World Bank Document

RESTRICTED TMrir ('(%DV TILL~~~~~~~~~~~ EAJST U

VOL. 3 Public Disclosure Authorized This report was prepared for use within the Bank and its affiliated organizations. They do not accept responsibility for its accuracy or completeness. The report may not be published nor may it be quoted as representing their views.

~ k ~' I ~TT,' T T%T1'1?-/V Tn"~T'r%TTf"?T'T%XT I XTT% ThT"I YTFT iN TRN A' .iAiONAL B.AIN' Pr. Rr' CN SiU.TI. LNAD LJ'nvn,L~..rLV1MJ.Nj

I TT r'TiNATI '1'T%kTI T 7rV 7Y TPT 1-kTrIT IrATIISC ON Public Disclosure Authorized

PROPOSALS FOR AN ACTION PROGRAM

EAST PAKISTAN AGRICULTURE AND WATER DEVELOPMENT

VOLUME III Public Disclosure Authorized THE WATER PROGRAM

July :17, 1970 Public Disclosure Authorized

Special Projects Department CURRENCY E14IVALENTS

US$ 1.00 = Rupees (Rs) 4.762 Rs 1.00 a US$ 0.21 Rs 1 million = US$ 210,000

MWEIfHTS AND MEASURS

1 Acre (ac) - 0.40q hsctAre (ha' 1 mile (mi) = 1.609 kil-meters (km) 1 square mile (sq mi) 640 nc = 259 ha 1 foot (ft) - 30.5 cen,timeters (cm) 1 long ton (ig ton) -1.016 metric tons (m ton) 3 1 ,.c4p.r'o+ (a-t = 1,234~cub,ic% meters (-m) 1 cubic foot per second (cusec) -0.028 .,./s

L"a.,U.L UAJ.LI,Vs

1 maund (md) 0o.0367 ig ton 1 maund = 82.2 lb

.NrIrALS AND ACRONihxS

EPADC - East Pakistan Agricultural Development Corporation EP4APDA - East Pakistan Water and Power Development Authority IRRI - International Rice Research Institute, Pnilippines PARD - Pakistan Academy for Rural Development (at Comilla) SIDA - Swedish International Development Authority TIP - Thana Irrigation Programme UNDP - United Nations Development Programme USAID - United States Agency for International Development USGS - United States Geological Survey

GLOSSARY aran - rice planted before the monsoon and harvested in November or December aus - rice planted in March and harvested in June or July boro - winter rice planted in October or November and harvested in Marnh or April

kharif - summer (wet) cropping season rabi - winter (dry) propping season EAST PAKISTAN AGRICULTURE AND WATER DEVELOPMENT

PROPOSALS FOR AN ACTION PROGRAM

VOLUME III --- WATER PROGRAM

Table of Contents Page CHlAPTERS 1

INTRODUCTION ...... 1 Geographical Setting ...... 1 Regions ...... 2 CHAPTER 2

HISTORY OF WATER DEVELOPMENT ...... , ,,e 7 SUMMARY ...... 7

ClHAPTER 3

INCREASING OUTPUT FROM EXISTING PROJECTS ...... 16 Croastl rm.hankmentsz 16 Phase I, Group A ...... 17 Phase I, Group B .oa..__ .._ ...... 17 Phase II ...... 18 Overall management assistance ...... Water management ...... 18 Organizational and institutional problems ... 19 Relationship with Southwest Region Water Stud- ...... +*...... Special Bank Mission ...... 20 Ganges-Kobadak Project (Kushtia Unit) ...... 21 Phase I ...... 22 Phase II ...... 26 Agricultural extension ...... 27 Conclusions on Phases I and II ...... 27

Chandpur Project ...... 28 CHAPTER 4

NEW PROJD'ECTS -- CUARDT-..DA1M73 ,~L.fl A~tA.l.t,±LkJflf.-l .lf,...... Low-lift Pumps ...... 31 Trub,ewell's ...... 3.) Existing tubewells ...... 37 Mu%C tube e l pr i ...... 3 EPWAPDA tubewell projects ...... 38

L%aL.qRat UkL t_UU-fubewell W.L. UCdeveop..ent V=.LUPLIILI...... 40 United States Geological Survey-assisted study ...... '...... 4; Indigenous methods ...... 41 Table of Contents Cont'd

Multi-purpose Projects - Feasibility Studies Available ...... 42 Dacca Southwest project ...... 44 The Karnafuli Irrigation project ...... 48 Muhuri project ...... 51 The Pabna project ...... 57 The Belkuchi project ...... 58 The Upper Kushiyara project ...... 59 Khowai project ...... 61 Sangu multi-purpose project ...... 62

New Multi-purpose Projects - Feasibility Studies Not Available ...... 64 The Dacca North project ...... 65 Barisal project ...... 66 Little Feni project ...... 68 The Comilla-Noakhali project ...... 68

Manpower Requirements ...... 70

Engineering ...... 70 Agriculture ...... 71

Agricultural Production ...... 71 Periodic Review of Projects ...... 74 Short-Range Proiects -- Summary ...... 74

CHAPTER 5

NEW PROJECTS -- LONG-RANGE ...... , 76 Flood Control Aspects ...... 76 Water Balance Asnects ... 77 Floating Pumping Stations ...... 82 Regions.83 Special Studies ...... 84 lG:nopg hbarraopc0nyld nintthwaet rgionn wteAar study ...... 84 Model of soutthwest regioon ...... 86 Flood models ...... 86

Brnhm.aniutrA-Canges-Meghna Complex - 87 Flood Consulting Board ...... 88 Basic Data ...... 8 Surface water hydrology ...... 89 Groundwater hydrology . .90 Aerial mapping ...... 90 Geotechnicall exploratio.,s ...... 9 Construction materials ...... 92 Sesmlogy ...... 93 Geomorphology and sedimentology ...... 93 Table of Contents Cont'd

TABLES

1. Cost Estimates 2. Project Areas 3. Incremental Agricultural Production from Water Development Projects

MAPS

1. Area Normally Flooded 2. EPWAPDA Water Projects

CHAPTER 1

INTRODUCTION

1.01 Chanter 6 of Volume I contained a descriDtion of the water devielon- ment program. In this Volume III the same program is described but in greater detail. Several of the larger Drolects (Pabna. Barisal. Little Feni and Comilla-Noakhali) have been divided into two phases and in Volume I the first phaAes of these nroiects were described. This volume containa further descriptions of these phases as well as particulars with respect to the later phases of these nrnioets= Tn addition- this voltump vivea information on long-range aspects of water development and on specific studies needed in connecrtin thprpzwith-

CpnerAnhirAl q'tttin0 l.O2 East Pakistan- Pakistan's eartern nrovince (area 55j000 square miles), is surrounded on the west, north and northeast by India, on the southeast by Burma, an on the snut-h hb fthe Bay f Bengal. WithIn the Province, the Brahmaputra-Jamuna, Ganges-Padma and Meghna Rivers join and flow Into the Bay of Bengal. Over the cent-r 4-es these three rivers have built the largest and one of the most fertile deltas in the world. Ranging generally- from sea level to 100 feet in elevation tdth ove half loer than 50 feet), this delta covers most of East Pakistan.

1.03 The drainage area of the three rivers covers 600,000 square miles bull' only 7.151IS percent,j~At tjlf.t.Af U * tAtof th4isUffltO lies,t within-A. flA.A EastLOOU Paki4sta-.nat la,. . TrherL CL Qreu.f_uL ULSSA. L.u- nately no storage sites within East Pakistan that could assist in flood or iow=flowA.A X U regulationL l La'tough OL. A UI L leaslbleaO±L -L~ sites-L.OL nnmay ex'stCA. L.W within LLI &.Lbea upstream, riparian countries. The long-range technical problem is further complicated by the great magr.itude of thte peal; fIloodu fIlow (5/C lo usc rao; ~L~OUU~tJJ.LLL~jJ~a%WU~L.L A LULU ILW _J IILL.L.L.LULon cusecs or AbuUI twice the historical peak flood of the Mississippi River) and the enormous of 2.4 sedlirment 'loadu th'e co`bnedI LUU r'ivers,JL which 'Ls about 4 bllonU.LJL.L.LMUULn.tons tonsI annually or greater than that of any other river system in the world. Th.e enor.u.ous transpo4ting power of the Ganges and Brahmaputra h.as resul:ed in major shifts of their channels in recent centuries.

1.04 Because East Pakistan depends on agriculture for economic survival, hler fl'ood ar.d drought prollems are particularly signr.ifican;. Zr. -le lar;t *5~I LLUIJUSIIU ULOU1 L. LFUL~U S pL LLU.LSL OL 1L JSL L L *LI L.iA WS few years, floods caused East Pakistan to import about one million tons of rice per year to fee Iler popuLatLon. IrL the f'uture, Loodu controL anu irrigation will be even more important because, at the present population growth rate of about 3 percent per year, East Paklistan will h ave 120 IIUI±IUIl people by 1985. They will consume twice as much rice as they do now.

1.05 The cultivable area of the Province (excluding rivers, streams, ___ _ _ .. __ .__.__ oo11) c4ll AL_- . _ l.-I__ _. . roauds, towns andu viln.;ges) Is aouuuL -22.5 million acres. AbJUL a LtLrU )I tnis area (from 8 to 10 million acres) is flooded annually on the average by spills from the main rivers to the land and from the .act that local runoff from the heavv monsoon rains is backed up by flood levels in the main rivers; see Map 1. Some flooding is also caused by tributary streams draining hilly areas either within India; or within Pakiatan but near the hnrder=

1.06 In the coastal belt, a zone covering 3 million acres is subiert to flooding from oceanic tides and tidal surges from typhoons. Depending on the time of the year and the flood flnwn in the rivers- the inundAtinna vary greatly as regards the degree of salinity.

1.07 A total of 17 million acres (two-thirds the cultivable land) is subiect from time to time to one of the fnreonino tvnpe of flnoding although the area flooded in any single year never has reached such a total. Because of flooinda 17 m1llinn tonn of rice were InAt hetween 1954 and I1S56 and 1.2 million tons were lost in 1962. The rice lost in 1962 had a value of $220 millionn

L08 Reaidaa airecat dlnnna tn ronns, floods hamner rice product-in ir. other ways. In deeply flooded lands (over 15 feet as in the Meghna valley) farmera cannot grow anv rice at all. Tn lande that are flonded from I tn 12 feet, they can grow only low-yielding braodcast aman rice. Where flood- inc ia frnm I to 3 feet, farmers grovw hiaher-yiliing aus. and transplanted aman rice but they must accept risks from occasional higher floods or from drouohtt nnd thav cannot' 1nelf, uase fe.rt'ilizers or imnroved seed effectively. The effect of flooding on land use is shown graphically on Map 1.

1.09 Rainfall in the Province averages 85 inches of which about 58 inches fall during the rainy season f.-o May to October. T' s monsoon brings rainfall averaging from 48 inches in the west to more than 100 inches in the H.imlay G foothllls. However even during the Monsoon period because of irregular rainfall there may at times be insufficient water for crops. Durng the dr; season, about 60 per.ent of te _ltivated land is left fallow because of insufficient water. Residual moisture is the b as4 for thse of th1e 510 O %1 .&cul:ivation AILI L t1*i1 %31 LIAC remalning 40 JLpercent LC with relatlvely 1ow- ~UC LCjU&K.L5 5 -VW U - LU4. . ~ L C .*w yielding pulses and cereals (mainly millets).

1.10 With water control (flood prevention, drainage and irrigation) .ouplJ.eU W.LL11 WVUdeL1 an8L.LLU±LULa.L LUpULU, tLAIC pLCs,IL aLcLLULLLL pLoUUcLtLoLI could be more than tripled.

Regions

1.11 One can divide East Pakistan into four roughly equal regions, based on hydrologic differences. ine Brahmaputra-Jamuna, Ganges-Padma, and Meghna Rivers form the boundaries of the four regions: the Southwest, the Northeast, the Southeast and tne Northwest. All four regions have severe winter droughts, and summer and early fall floods. In the Northwest, droughts are more serious than floods whereas in the Northeast the reverse is true. Both droughts and floods are serious in the Southeast and South- west regions; the floods in these cases are mainly tidal in nature. 1.12 The Southwest Region, which covers almost 10 million acres, consists of Kushtia, Jessore, Faridpur, Knuina, Barisal and Patuakhaii Districts. Each year, cyclones produce high tides in rivers emptying into the Bay of Bengal; tidal surges from typhoons sometimes reach up to 25 feet above sea level.

1.13 In the past, tidal floods especially hampered the agricultural otutput in an area of 3 million acres approximately contiguous with the laitter three Districts and resulted in extensive salinization of the soil. Farmters planted transplanted aman rice in the wet season after the rains had partially leached the residual salts from the earth. In the winter, the drought: and soil's excess salinity prevented farmers from growing any crops. For these reasons, only a single crop had been grown in these Districts (see Map 1). During the past 10 years considerable progress in the construction of embankments (see Coastal Embankment project described in Chapter 3) his protected an area of about 1 million acres from saline inundation with resulting increased yields in the transplanted aman crops. Within this zone, further increase in yields as well as the introduction of dry season crops by means of irrigation, to the extent that water can be made available, will come as a later stage. The remaining area of 2 million acres still suffers from tidal inundation but with the important difference that the flooding is from fresh rather than saline water. It seems likely therefore that dry- season irrigation should in this case come first and flood prevention as a later stage (see Chapter 3).

1.14 Although river and rain flooding are not as serious in the southwest as tidal inundation, they do harm much of this region's agriculture especially in the Faridpur and parts of the Barisal Districts. Distribu- taries from the Padma flow, such as the Arial Khan, periodically flood the latter two Districts. Because of these floods, Faridpur's yield of rice is only 75 percent of the East Pakistan average. Consequently, Faridpur's large population consumes more rice than her farmers can produce and the Government has had to import rice to feed the local population.

1.15 Drought rather than flood is a problem in Kushtia and Jessore Districts where ground elevations are higher. Cropping intensities average about 130 percent with considerable areas in rabi crops (pulses and sugarcane) that utilize (in late fall and early winter) residual moisture stored in the soil. Aus rice is another widespread crop in these Districts but yields are low due to dry conditions in the early part of the growing season. The Ganges-Kobadak irrigation project (Kushtia Unit) covers 0.35 million acres in Kushtia and the northern part of Jessore District (see Chapter 3 for detail). The Ganges-Kobadak development at its ultimate stage would cover a total area 2.6 million acres. The Government's intention has been to base this irrigation almost entirely on diversion of water from the Ganges River. The Government has in fact proceeded with design of a major barrage one of whose functions would be to facilitate such diversion. However, whether such an ultimate plan is technically and economically feasible is not yet clear. For one thing, an agreement with Indla would probably be needed - 4 - with respect to water releases at the Farakka barrage and there are many other tArhniAl onmnlplexities to h workedA ot wi4th res- ac to oher effects of upstream diversions (includes increased saline water intrusion, reduced na-vivat4nn tipnths nndi rpieduced v4eldsc fro^m f4cishng) ."A wi4th re-ect to possible counter-measures (such as use of return flow in downstream areas from upntream 4rr4 gatlonn, poss4ble closure of tidal estuaries and development of groundwater). (For further detail see Chapter 5 and especially the nrnnopod spei-4al st4udies of the Ganges barrage, the Southwest region and the Brahmaputra-Ganges-Meghna Complex.)

1.16 The Southeast region (gross area 7.7 million acres) includes the Dist:ricts of ro--411a, Noakuhal Chtt_ogO,4 n th-en-,ltaon ilI Tracs s .. 1 1 U-. -- L - L ~~6 & cIL&u LJit %.LILL" YVg LAr, El LL.LA. L racts. Tidal inundation is one of this region Is major water problem. Much of this inundation ia saline but some of this has pre-vented…een by te Coastal ~~I.X L.L L u 1U- L LV.LL II ~ p U LLILIX Embankment project (Chapter 3) which extends in part to the Southeast region.

1.17 River and rain flooding and droughts also have harmful effects. ,,e MeULna an' U ivers flouo LarLg sectio-11 VI LfNoaK118ha1 anu aImoSL all of Comilla. The Karnafuli, although in parts under control, and the Sangu River flood parts of the C.f-ttagong andA t1- ('11. TrIl t D The Southeast produces a large crop of broadcast aman in the summer and early fall when floo(As in some places reachn dept"hs of I) to 15 feet. secause of droughts, farmer only produce a small winter rice crop and some rabi crops (from residL'Uua'l soil UmoLsturel). 7Iey also grow a suIaLL crop ofL transplante aman in less flooded areas.

1.18 The high elevations in the Chittagong Hill Tracts and parts of the C`h1ittagorng i'stricts prevent floods from occurr'ing, aithough winter droughts are still a problem. The soil here is sandier and has a different composition fLrom the rest ofL East PLKim2istan, thus entabii'ng f'armers to grow citrus fruits and tea on the hills and cotton on the lower slopes.

1.19 The Northeast region (gross area 9.2 million acres) covers the following districts: Mymensingh, Sylhet, Dacca and the northern tip oi Comilla. This region's topography, climate and rivers all contribute to East Pakistan's worst flood problem. Tne region includes the Syinet Basin through which the Brahmaputra River flowed before it changed its course during the latter part or tne eignheentn century. mucn or tne bylnet Basin lies between elevations 10 to 20 feet as a result of 30-40 feet of subsidence that has occurred over the past several hundred years. Because of the plain's low relief, channels and basins are not deep enough to drain the Brahmaputra-Jamuna and Meghna Rivers" overflow during the monsoon season. Nor can they drain the monsoon rains averaging between 220 inches in northeastern Syihet to 70 inches in Dacca.

1.20 Floods and droughts slow down agricultural production in this region. Meghna River Valley farmers cannot grow any variety of rice because floods rise to 20 feet here. Floods damage crops in Sylhet and Mymensingh. In Dacca, agricultural workers can double crop only 40 percent of their land. Dacca's large population consumes much more rice than the flood burdened District can produce. In the last few years, the Pakistan government imported rice to satisfy this population's needs. 1.21 Farmers in these zones adapt their agriculture to drought and ,1wod cond't'Lors. hTere floodling is irfrequent, far-ers gLow Jute, aus, and transplanted aman during the monsoon season. Where floods do not exceed fifteen feet, they plant broadcast wminr. On the molst low lands, farmers grow boro in the winter. But in other areas, the drought prevents a borco crop. SyiLhet's hils5 are not flooded atud thuue hil.: l soils lavea diUffL:[lerie: composition, thus permitting the East Pakistanis to grow tea and citrus fruits.

i.22 Tne ultimate solution of the Northeast region's water problems involves serious complexities. Aside from the areas of the Dacca Southwest, Dacca North projects and several small tributary projects in the extreme Northeast (total area about 0.7 million acres) plus some areas totalling perhaps 0.1 million acres suitable in the near future for groundwater development, the best method of proceeding in the rest of the region is uncertain at this time. A possibility that has been considered for many years but needs a great deal of further study is the diversion of a portion of the flow of the Brahmaputra River into its former channel (designated "Old Brahmaputra River" on Maps 1 and 2) and thence to the Sylhet depression; see Chapter 5 on description of the "Brahmaputra-Ganges-Meghna Complex".

1.23 The Northwest Region (gross area 8.5 million acres) includes the districts of Dinajpur, Rangpur, Bogra, Rajshahi, and Pabna. Drought is this region's most serious problem since the monsoon season is shorter here than elsewhere in the Province. The drought lasts seven months, and prevents farmers from growing a winter crop of rice except in a small portion of the lower-lying lands.

1.24 Floods are also a problem in the Northwest. While the right bank of the Brahmaputra-Jamuna has now been largely protected by an embankment (completed in 1968) the Teesta and Atrai Rivers continue to overflow. A large section of southeastern Pabna District suffers flooding from the Ganges each year (see Maps 1 and 2). As a result rice production is low and in the last few years the Government has imported rice to satisfy Pabna's needs.

1.25 Northwesterners adapt their farming to flood and drought conditions. In higher lands they grow only transplanted aman as there is not enough moisture to start an aus crop before the monsoons. In intermediate lands farmers grow aus or jute alone or double-cropped with aman or sugarcane. In low-lying lands they plant a mixture of broadcast aman and aus or broadcasL aman alone (where flooding is deep).

1.26 As is generally found to be true in East Pakistan, the zones of flooding are also the zones where surface water is available for irrigation. In the Northwest region this applies particularly to the eastern part of Pabna District. Elsewhere in the region, surface water is scarce and irrigation will have to be based largely on groundwater. Fortunately, the - 6 - outlook is favorable in a considerable part of the Province. A tubewell proiect (Thakurgaon) for the irrigation of a concentrated zone of 70.000 acres was completed in 1967. The Agricultural Development Corporation intendR soon to proceed with a pro1ect for 3.000 tubewells at dispersed locations, mostly in the Northwest region; see Chapter 4. -7-

CHAPTER 2

HISTORY OF WATER DEVELOPMENT

SUMMARY

i. Imvortant construction efforts have been successfully carried out in East Pakistan but mainly with respect to single-purpose flood prevention proiects. The achievement to date in this type of prolect amounts to about 1,000,000 acres protected against tidal flooding and a similar area protected against river flooding. Construction progress, while slow at first, has improved in recent years. The use of labor-intensive ("head basket") methods for the extensive earthworks involved are well suited to the local conditions of climate and labor supply.

ii. The execution of multi-purpose projects, involving three functLons -- irrigation and drainage as well as flood orevention -- began about 15 vears ago but has not made the rate of progress that had been hoped for sinre ihe Arh1evpment tn date amounts to only about 60;000 acres enJoying benefits from all three functions. The disappointing results obtained tlhus far from this type of pro4ect hba heen dtip to a ronmhinAtinn of fartors among which probably the most important has been a lack of understanding of the the particular requirements of surh pronects under East Pakistan conditions. This understanding has however made important progress in the past few years and there is now keen reanliatinn t-hat two% key InoredIents for success are: (a) a well integrated project organization capable of dealing with the farmers (the eoatahliQhment of asch an organia7ntion involvas coope- tion among several agencies including EPWAPDA, the Department of Agriculture, the A^nC and the BD & LG Departm-ent) and (b) adoption of n methoA of ir4lga- tion distribution suitable to the conditions of small, fragmented landholdings that prevaIl. Several prniects aggregating 450nOOO acres have hben prepared on the basis of these concepts and either are under construction or at the Qstag of final Pnpineerino des4gn_ iii. Exr.4noin with the l1w-14ft npmmn nrnorram of Anlr (nnavoar,nh 9 1AL indicates that East Pakistan farmers do in fact respond quickly when irrigation water is provided 4n -he dry sean.n Th. aran irrlgated (mnstly bhiro-i rice during the winter months) has grown from nil in 1962 to about 200,000 acres n4 10A6/97 anA 700 nnn acrea in 1Q9Q670. Th. areas irrigated by this means to date have been located along perennial streams where conditions have been relat4vely favorable. The stage will soon be reached when medlu-Scale civil works (i.e. EPWAPDA-type projects) will be needed so as to increase water availability and thereby greatly extend the area that can be irrigated by low-lift pumps. Such civil works will at the same time give major benefits through flood preveration., drainage, navigation ar.d im.proved fs culture. - 8 -

iv. The use of tubewells has resulted thus far in the irrigation of about 100,000 acres. A project for an additional 150,000 acres is just starting. It is noteworthy nevertheless that the exploitation of groundwater for irrigation has barely started in East Pakistan. As discussed later in this volume, as further knowledge and experience are gained- ground- water development is likely to grow in relative importance in the years ahead.

2.01 Small-scale irrigation in the form of nrimitlve lifting devires has a long history in East Pakistan in view of the lack of rainfall during the winter months. As these deviceR are inefficient and can lift nnlv qmall quantities of water, the area cultivated in winter through such means has however remained small The use of diesel-driven pnmn%s obtaining water from perennial streams, began in the early 50's and by 1955, 40 such pumps were ofnerAtina The RAat Pakistan Agrie-ilt,tral Devealnnmant Conrporatinn (ADC) was established in.1961. While its primary functions are the procurement and distribution to farmers of aRricultural inputs (seeds. fertilizers, pesticides and agricultural machinery), ADC has also a highly imnortant secondna-y f2nc-tinn; nAmely the nrnmntion of the develpnnmnt- of small-scale irrigation by means of low-lift pumps and tubewells. ADC has hben A mAin supp0I4r and niarnniyav cin-a lQA7 nf orvtable, dseaal-dr4ven low-lift pumps which generally have a capacity of 2 cusecs and are almost entirely uisedRonly dwrino t-he hnorn asenn being r.^xuled fr-om t-he field at the end of this season. By 1966/67 a total of about 4,000 such pumps were supnlied by Anfc In the 1969/70 season th4i increaseadA toa !80 pnmp,anA the Government has plans to increase this up to 40,000 for the irrigation of abt..-t 1 1; 114ion acres at scattered locations througout the Province. This is about the upper limit of area irrigable by low-lift pumps, unless a-asisted by lar1 ar scalae civil works (see belo r..d in Ch.apter 4).

2 .02 Sir.ce a tprc low-lIft pup of 2-cVsec capacity seres up to about 50 acres and since, with the small landholdings prevalent in East Pakistan this area cma involve as many as 50 fa = rs, a key element iz the organization of groups of farmers for the distribution and utilization of irr i::igtionater.~igaa-4on WUu 'Po~ assistaW l in -thisIL Ji5o ect-ive a-L leastJtOI.llOJ i.-sofarL, asthe1LL.V low=L. lift pumps are concerned, the Government a few years ago established the so-called T.k.Lana Irrigatio Pl'a. P(T1P 1 I un.der hichl th res1pr ibi-t for aL..a.LL~UJililo LLL~UL.IJ LJdll ~L.LJ. ±.Lj uIJ WUL Itl Lilt L=O WLAOJ.U.LA.LLL LUL forming pump groups, for providing and paying pump operators, for providing uel. mu' fLor coLilectLIng p-up rentnal charges ha 5 Ueen thile 'LunctiLon of tLile Basic Democracies and Local Government (BD & LG) Department utilizing the existlngL uUnion anu 1111ar&aCYoaJucils. AWDCUs sole -responsibility 'LrsthLus Lhe provision and maintenance of the pumps.

l/ See: Volume I, Chapter VII; also paras 4.05-4.09 below. - 9 -

2.03 A1)C in recent years has also become involved in the construction and nnperation of tubewells and has constructed 800 wells of 1.5 to 2-cunic capacity during the past 3 years. These are at dispersed locations rather than in a en-ent-rared zone auch aq those of EPWAPDA (deostrihed hblow!- As for the low-lift pumps, distribution to farmers' field channels is being hnneilA ,under trhe TTP- TDA in nrnopaeina a nradlt for 3,000 ADM t,,hbwella to be constructed during the next few years; see Chapter 4 for further

2.04 Tar--e-scale water development in Et Pistan began. in the early 1950's. At that time after several years of studies, a team of UN experts propoAed the Glangess-KOb-adak proJ4ecta coverAing 1 .9 mi0. lion.4 acresr irn t.he Districts of Kushtia, Jessore and Khulna. 1/ Continued assistance was 4 nr.8_.44AA 1.. i-ho CsaA _a A An n,.i?.,ra flrnan4 ..a*n AS i-l. TlT C{_ i-his ___ ject, the Kushtia Unit of which wasA then started with aid in the form of foreigrn excha.nge pr.ov 4Aded- r.ada and the Un4ited4 States. In '9554, 1955 and 1956 severe flood damages occurred and the Government of Pakistan again

sought the assltan.ce of the TT4i-eA Nati-ons. A tea. ofAexperts headed by J.A. Krug, former Secretary of Interior of the U.S.A., visited East Pakistan

*in ELCate 5%6JUL . G- Lqnelr * kno&WL as t4he Kru Missi oL n tJL X fiCprU J.7.aJ / recommended the creation of an autonomous corporation to deal with water and e'lectriLc po-wer development. Such a corporation, 'Known.as the Eoa °al; Water and Power Development Authority (EPWAPDA) was established in 1959.

2.05 Since EPWAPDA's establishment in 1959, six agricultural water

pro'ects hIlave h'1adUpr,Lmary LmportL. ce, as 'Lolluws

Complete proects

Brahmaputra Right EMbankment Northern Tubewells (Thakurgaon) Dacca-Demra

Partiallycomnpleted projLcts

Coastal Embankments Ganges-Kobadak, Kushtia Unit Cnandpur

The main emphasis thus far has been on single-purpose flood prevention projects, especially in the coastal area.

2.06 Among the six existing projects listed above, the coastal embankment project has constituted by far the largest effort. A sum of $200 million (total equivalent) has been invested since the project started in 1959. The principal source of financing has been rupees generated by the US PL-480

l/ Later increased to 2.6 million acres (ultimate); see Chapter 1. - 10 - program. The project will ultimately provide protection against inundation from nepanle tides in an area nf I mil4Unn A-rrp giih-dAiridpd into 102 cnpn- rate polders. Inundation is of two main types: saline and non-saline, with varlousq radattonn thereof.

2-07 Phase I of the pronect consists of 86 polders with nn aggregate area of 2.0 million acres. To date 61 polders have been completed providing nrnrprt-inn rn An arpe of 1 0 million acres; the remaining 25 polders in Phase I are partially completed. Construction has consisted primarily of earth emhbnkmen.ts. Th .ese have an average height of about 12 feet and their aggregate length is over 2000 miles. Numerous sluices have been provided through the eab F.ts provi outletas for 4nterior ra-i4nage

- is'. -. I iL 4t'O * h.L LL---. L, 2.08 Alth1ough, the pro4ect h.s proviAe-A otntbnft 6 truhei mination of saline flooding, there have also been cases of negative benefits since farmuers ir. somm. of th1,e polAers had previously beer. accustor,.ed lo uslng non-saline flood waters for irrigation. As a next stage of development, water muaagement ir. th--e irterior of eachl. of the poldUers, coup'ledU with agricultural inputs, would provide sizable benefits at relatively low cost. A program for such development is described in Chapter 3.

2.0i Th.e second mALost important enAeavor bLy EWIADA has been the GK- Kushtia project in which the total sum invested since 1959 has been $60 mUI4.JIon toLs-al CeqUiLV alentI . Fi UancLini LIas UCCeL miain'LIy *.roIU UL.l resourc-sC but with substantial foreign assistance from Canada and the United States. The ,.'.. p-rp- . -o-4nth 4n i- .4t Ap 4d -4rr-4gati .d-A ra-i4 age -A e14li4 VLC[UutjJLiCC& US -In FJLWJC~- . LU jJL'V4.UC XLAL A6"UI.LWStJll& OU"iLLCL6a&u05 afLS.- C~lUU nate flooding in a gross area of 350,000 acres (120,000 acres in Phase I and 230,000 acres 4r a. eC II -= onLy IPLhase I is In operation at present). Although the project is providing benefits, these have fallen short of ex- pectations since irrigation i8 beLng provLU'ided for only 40 ,000 acres and even this occurs only at the end of and immediately following the monsoon season. T.hi'si8eca-use the pro'ect has suffjeredu frorm severali ser'Lous technical difficulties and because the type of irrigation distribution (conventional gravlty systemj auopteu imay not be 8u'ited to East Pakistan condtuiLtLons (very small and badly fragmented landholdings). On the other hand, GK-Kushtia can still provide substantial benefits. Tne serious technical difficulties mentioned are in fact gradually being corrected; a stepped-up effort is needed however as discussed in Chapter 3.

2.10 Tne Brahmaputra Right Embankment project was constructed in the years 1964-68. This project for which IDA provided a credit of $5 million consists of an embankment 135 miles long and 12 feet high (average) and appurtenant structures consisting of drainage sluices. Protection against floods from the Brahmaputra River have been provided for 580,000 acres and important benefits are being realized, mainly through higher yields from transplanted aman rice rather than the lower yielding broadcast aman previously grown (see para. 5.04). As a next stage the protected area will require irrigation facilities probably by means of tubewells; see para. 4.30. - 11 -

2.11 IDA has so far financed three water development projects in East Pakistan with a total of US$24.25 million of credits. Besides the Brahmaputra Embankment project these have included the Dacca-Demra and Chandpur nrnioetR. An IDA credit of US$1.0 million for Dacca-Demra was provided in 1961. The project provides partial flood protection and irri- cgation in ansrea nf 15,000 aCreS. It wan suhbtantaA1lv rncmpnlted in lQ9i7 and benefits are now being obtained.

2.12 The Northern Tubewell (Thakurgaon) project was substantially completed in 1967. It consists of 365 electrical!--driven relativel, lare- …---~~~~~~~~ .7 -'-o-- capacitv (3 cusec) tubewells that will ultimately serve an irrigated area 4 4 of about 71,000 acres. T.h.e pro ect cost was R-*. 155lo411 ^. ndA L- aaa4a#i A by two German loans totalling DM 22.2 million. As part of the project, a brikck-li4nedA cnal generally several h.ndred feet long has been provi4AA leading from each well. Since completion, additional expenditures for lining of fielA c-nanels th f.fat.fromtSU the brl.k-lne -.-- I (the region is sandy) are being found necessary. Even after full development, th1Je projec. will be costly whether uncasured ir. terLs of coat per well1 or per acre irrigated. The long and narrow configuration of the project area and the dispersal of the well sites, coupled with. the Aesigre of providi;;g %A I 1 .AU . ~ 30.. J. ~S.f ~ ~ . 0.3.L Jtr .t AL.l LI AL 5.A j./LtJ social benefits through rural electrification are the apparent explanatLons

Oj. tlleof hl,l111igh11te 0Uost.EPWDTADyA Lt £."L " fIA LILt isL.L n-&L%JW conucin1L5.UL.LAU6.LI&J stu-eL.U%UL=O LLIC&V..h;cud-edt %.UU.LU .LCu L'U a more rational use of the electric transmission system through the addition o1. tiVore We.L' s. ILrLrLigaLton nAL -ImprLoVedU a gricLULtur arL Ue'Lng UeVeloUpe through cooperation with the Pakistan Academy for Rural Development at C..omlla which has1 poLUViLUU rLLextension director sin.ce iUd-19766 and the crops that have been developed in the project area are generally good.

2.13 In 1963 IDA granted a credit of US$9.0 million for the Chandpur project, a combined irrigation and flood protection scheme covering 127,000 1/ acres. The irrigation features were dropped in late 1965 and the credit reduced to US$5.25 million. Work proceeded on tne flood controiL features but these too were halted early in 1967 and the remaining proceeds oLfthne credit were useud to engage nW corsultarUtLs kin i19O8) LV asULUL zLn the preparation of a revised project which would restore irrigation as a primary feature of the project but on a substantially modified Dasis as compared with the original project. Whereas, under the original project, irrigation distribution was to have been accomplished by means of canals elevated above the terrain (which involve high construction costs and extensive land acquisition), the revised project would make use of natural drainage channels requiring the acquisition of very much less land. W'ater would be lifted from the drainage channels by means of portable low-lift pumps. Finally, flooding within the project area would be eliminated gradually over a period of several years to allow farmers time to adjust to the new cropping conditions provided by water control (see para. 5.04). in April 197u !DA granted a credit of US$13.0 million to assist in financing completion of the Chandpur project, as revised.

1/ Later changed to 140,000 acres gross (/5,000 acres net). - 12 -

2.14 The idea of using low-lift pumps for irrigation distribution within polder-type projects such as Chandpur is a recent development since, as indicated above, low-lift pumps have been used on a significant scale in E-at Pakistan or.l during the past four or f years.0Ve "Ue layout of f ieldC channels for water distribution within a typical 50-acre area served by a low-lift pup, while impeded by the fragmtented holdings -.d small p'ots, is far less complex and less land consuming than what would be required for a conventional gravity system. ,/ PriLmaril.y ecause of quick luenefits obtaLn- able during the dry season, low-lift pump irrigation is catching on quickly 4n East PziD1stan. Without sizab.le engineering work-s 'or primary pumping - u ~NELJ L.L * VL 9 LYLL.L L .LU± JLLiI L L rWL,13 L~ upn from the larger rivers however, the scope of low-lift pumping using peren- niaL low flows of the smaller rLvers xs limLted (the present estimate is about 1,500,000 acres). Besides simplifying land acquisition, the use of low-life pu-m-p8 aLiU ntaturaL chlannieLs i1or water diseribution nhas several other important advantages.

One advantage concerns conservation of water supply. Take Cnandpur as an example. An embankment will completely surround the project area thus forming a "polder". Water will bL'e iLntrodLuceUd iLnto thLe poldUer eithLer by gravity -wilen river stages are high) or by means of a large primary pumping station (when river stages are !ow). Water thus introuduceu will be distributed throughout the polder via the natural drainage channels from which it will be relifted by the low- lift pumps. The waste water (water not actually consumed by crops) will return to the drainage channels but will never leave the polder. Water will thus be recirculated within a "closed system". (Waste water is very high -- as much as 75 percent of the water that is actually supplied to farmers' fields. Tne high wastage results from the fact that it is impracticable to install a rational distribution system because of the unusually small and fragmented holdings. In the Chandpur project area the average holding is 2 acres and this is fragmented into as many as i0 pieces.) The "closed system" differs from an "open system" of the gravity type (such as that of Ganges Kobadak project) wherein waste water leaves the project area.

1/ In the long run, once the value of irrigation is fully understood by the farmers they will gradually realize that not only can they irrigate and cultivate more efficiently if their holdings are consolidated but also that the system of water distribution can be simplified. Such simplifications would include: (a) more rational layout of field channels, thus reducing maintenance costs and water wastage; (b) replacement of some of the portable pumps by fixed installations of larger capacity serving a larger area, with resulting economies of scale (to accomplish this might require some gravity canals). - 13 -

In the "closed system" the low-lift pumps recirculate waste water so that only "make-up" water needs to be supplied. The primary pumping station serving the closed system does not need to supnnlv waste water as does one serving an open system. The make-up water equals in volume the crop requirements ("consump-- t-lvp uisse") nlun minor lnnsse sueh an evapnratinn frnm the water surfaces of the drainage channels, seepage at the periphery of the nnplder And 1nekAQe loasse (at navigatrin lrocks). While precise data are not now available it is certain that the aggregAta nf these minor losses is far loan than thea onnatino wastes associated with an "open system".

A second advantage is that farmers pay operating costs of the

will be heavy at first but will be gradually eliminated. The operating cost will thus act as an incer.tive to rce water consumption.

A third advantage is that the closed system provides maximum

1. .cA.LU-L .5. Ly&LI0 .5 L &6 At. uOg L.L a w.U: VaL L=Cy VUL %.;LUJP O . L can be expected that at least in the beginning high water- requirement crops sauch arS Lboro rice will ble introduced. Excessively large canals would be required with an open gravity systm.heras,with1 a closedI system,, the natural channels s ys~ temI WLI4.i~ Lil L±U~ ui L~il LLI LIIULL LIL iL± (which supply only make-up water) can more easily supply the

LIL5IItOhl g'-water-requ'rem,en' L~(JL IIL.LLUf crops.

2.15 Sir,ce its formation, ELrwrDA has devoted much elfort to project preparation. Feasibility studies have been completed for more than 40 projects and' are uniuerway Lor abuout anothler 10 projects.

2.1 Following a request made inr May 1968 by UVP to the Bank, the East Pakistan Agriculture and Water Development Bank Group conducted a

review of 32 proJects tiatL "Er"t%VI, iLLdUiLcaU shIUU.LU bU cs11bdeUCLeU LUL inclusion in the Fourth Five Year Plan (1970/75). This review involved an examilaLtlon of feasibiltLy reportsLor the 32 projects supplemented by field trips to most of the project sites and discussions with EPWAPDA and its project consultants. l/ ine review, wnicn was completed in August 1.969, confirmed an earlier judgement that had been reached by the Bank Group 2/ namely that, while the feasibility studies prepared by or for EPWAPDA up to that time contained much valuable information, there were certain defi- ciencies needing correction. Besides tne problem of irrigation distribution,

i/ Summarized in report entitled "Observations on Water Development Projects, Fourth Five Year Plan", Octoher 23, 1969.

2/ Report of November 1967 Mission, May 21, 1968. - 14 - two matters that generally had not received adequate treatment were: economic analysis and project organization especially as regards agricultural inputs. (For more detail see Chapter 4 under Multi-purpose Projects.)

2.17 An opportunity to solve these questions in an integrated manner arose in 1968 in connection with the Chandpur project (para. 2.13). The consultants engaged at that time worked in close cooperation with the East Pakistan Bank Group in revising the project as regards irrigation distribution and organizational aspects. In connect with the latter, the consultants, following discussion with agencies other than EPWAPDA (e.g. the Agriculture Department and the ADC), recommended an organization that could deal with oroiect implementation at the farmers' level. A revised feasibility study along these lines, which was completed in September 1969, formed the basis for an appraisal by IDA of the second construction credit for this project in the amount of $13 million (see para. 2.13). This credit will enable comnletion of construction (bv EPWAPDA) as well as the establishment of a project implementing organization which the Government has decided will be the resnonsibilitv of the Department of Agriculture. The organization will include personnel for agricultural extension; plant protection: installation, operation and maintenance of pumps; and assistance to cooperatives. All such personnel will be deputed from existing Govern- ment agenciePs and w411 be under the control and supervision of a Proiect Director appointed by the Secretary of Agriculture. (For further detail on the rhAndnuir nrnioet see Chanter 3-)

2.18 A qernnd case arose in mid-1969 in connection with the Dacca Southwest project (net irrigable area 270,000 acres). A feasibility study for this proJect, which covers an area about three times that of Chandpur, was completed in March 1968. The Bank after detailed evaluation of the available infora-tion concrlutdmed that Dacrca Southwest was basically a technically and economically sound project, but that further work on organizational aspects, on the repayment prnhlem a,nd nn certain technical features (especially as regards design of the flood embankments) would be needed prior to a"r,.4anl nf a redirt fnr rnnital investment The Bank therefore proceeded to appraisal of an engineering credit which would llow the project to proceed to fina1 en4ineerlnraw4hilr gncu,rrneOtv rF- vising the March 1968 feasibility study so as to eliminate its defici- ecies ZA_ TnA - -.-._ A ., 4- n- cemQ er a f99ne amou,nt of £L .LJ C- S . -_E Cn v_ S £ A -w -6 $800,000 for this purpose on the basis of which construction could begin .JanuaryJar,uary .I I971J.1 providedjJULUU that6LIOa.. fina.ncinIJ*UAS.A.lt canS4l lbe ar.r.-nged.

.A4.1-e~..4 4." AA i 2.19efl ~ tA s'.lar....- l. IAer,gineeringTN A crei in the wamount of $,0,0, ~9LLnA covering the Karnafuli and Muhuri projects was granted in June 1970. The net irrigable areas are 6 7,00 ar.d4,000 acres respectively.

2.20 Besides the three construction cred;ts (fparas. 2.1, 2.11 -.d 2.13), the engineering credits for the Dacca Southwest project (para. 2.18) and the Karnafuli and Muhuri projects (para. 2.19), and the formation of the East Pakistan Bank Group (para. 2.16), there has been one further important item of assistance to EPWAPDA by IBRD/ID'A. mnis has been the - 15 -

financing to the extent of US$2.0 million which together with $1,947,300 from the United Nations Development Program will pay for the foreign exchange cost of general consultants to EPWAPDA during an initial two-year period starting in early 1969. General consultants to EPWAPDA during the previous eight-year period (1961-68) had been financed by USAID. Advice on management aspects 1/ had been provided by a group of three specialists from the Department of Water Resources, State of California; this assistance, from 1965 to 1969, was also financed by USAID. The scope of work for the new general consultants, which is similar to that of the former one, includes assistance to EPWAPDA with respect to basic policy, plans and programs related to water and power development; also the development of an efficient organization to plan, finance, construct, operate and maintain water and power development projects. The General Consultants' assistance is divided into three parts: services to the Water Wing of EPWAPDA, services to the Power Wing and services to all parts of EPWAPDA. in the field of management (the scope of work thus covers what had formerly been the concern of the California team). A highly significant part of the General Consultants' work is in the field of project preparation and final engineering. The General Consultants do not do this directly but rather through separate project consultants whose work the General Consul- tants assist in monitoring. The General Consultants also assist EPWAPDA. in the drawing up of specific terms of reference for the project consulting firms in their selection and in review of aualifications of the specialist personnel actually assigned.

1/ Tnne1 iedp development of program And f-niidrt-t i-nnt-i-rla. imfrntvrme"!:s in accounting procedures and inventory controls; and improvements i n organizational 8tiucture nand in -irso.n.nel development Jncu-ding st.:ff training. - lb -

CHAPTER 3

TNCREASING OUTPUTT FPROM. IISTc'INC( PROJT,TC

Coastal Embankments

3.01 Among the six principal existing projects, the Coastal Embank-

_l A 0 S_ A A 1 _vA*F_ { tl A {()4 1 I 1 1; I __1{ III[LL3 .LD Uy Lar LLALe largest. rt SUM1I tGL M.'V. -LX.L.Lon \LULAotl eq{UivadALen) has been invested up to now. The project hlas thus far concentrated on flood protection against oceanic tid'es. Dlespixte benefits obtained fror,- L £IJIU VL L±J L L. LL. L U~* u p t UI~L.Li (l. dLLU 1L01 such protection, much larger benefits could be obtained through interior water uadnagerierit andU provLsiLon of agriLcu.LturaL 'iLnuts. Water management would have several facets including: (a) introduction into the polders

o0 nlon-sal'lne riveL WaLteL dIIU 1 Lts uiLsLtriuLon uv various means; (b) conservation of natural rainfall mainly through adoption by farmers of proper cultural practices; and (c) development or deep as weli as shallow groundwater to the extent available. Preliminary information is available with respect to thLie extert of areas thilat coulu be developed by the first means -- it might cover 1.5 million acres or about half the total area of the coastal embankments. Closure of some of the tidai estuaries might extend this area considerablv but the feasibility of such schemes can be determined only arter detailed study (see water study of the Southwest region, described in Chapter 5). The extent to which the other two methods are applicable is not known at tue present time.

3.02 Because of its size (3,000,000 acres of which 1,000,000 acres now have fiood protection with the embankments for an additionai 500,000 acres partially completed), remoteness, and various technical and socio- logic complexities, the Coastal Embankment area will require special treatment and effort. Such special effort appears well justified in view of the large potential of the area and the relativeiv low capital investments required. The speed with which the potential of the area can be realized will depend primarily on the resolution ot organizational and institutional problems. The Bank is prepared to send to Pakistan a special mission to assist the Government in formulating, in the first instance, the institutional framework needed. Concurrently, "pilot polders" for demonstrating the value of water management and agricultural inputs can be advanced as well as technical studies for their gradual extension to a major part of the entire coastal embankment area.

3.03 Phase I of the Coastal Embankment Project consists of 86 polders enclosing an aggregate area of about 2 million acres. As of Julv 1970, 61 polders will have been completed, enclosing about 1 million acres. Of the remaining 25 polders, some construction has been started on 17. "Group A" designates the Phase I polders which are either completed or started (78 polders covering 1.5 million acres) while "Group B" - 17 -

designates thie 8 remaining polders (covering 0.5 million acres) not yet started. The Group A polders are scheduled for completion by EPWAPDA in 1970-71 and 1971-72 at a cost of $40 million. These polders have had nublic acceptance in varving degrees.

3.04 The Group A polders need additional project planning and water management to mitigate the loss of beneficial fresh river overflow in some areas and to increase the water supply and management in all the polders. The Group B polders have been deferred due to unacceptance by the public primarily because they would interfere with freshwater inflows during October and because intensive polder planning will need to be completed before construction can properly resume. For these reasons construction expenditures will be minimal in Group B for the next three vears with emphasis on proiect plannin2 as is indicated in Table 1.

3.05 Phase II construction would be deferred to nermit nolder by polder project planning and management studies to be completed with a construction start scheduled for 1972-73. An overall manaopmpnt ass1 st- ance program, including all facets of implementation from management of water s,nnlv- oneration of nhvyiral works; anrirultiirnl Pyt-pnsion t1hrotigh advisory services at all levels has been programmed for the entire Coastrl Fmhankment nroniert as indirated in Tahle 1. The mAnaoement assistance program, together with the project planning studies now proposed, ion,lld he- direrted toward cnluing the many problems of pro4ect14plemIn- tatton including the development of a viable overall organization to direct the proj,ect.

f it", Th fnl 1 va

3.07 Phase I, Group A: This designates the 61 completed Phase I

polders(r . 1140--n acres) andIthle l17 L-LICat hLave been started bCILut not completed (area 0.5 million acres). Generally these polders have been. 4 zl o,nes ofF cM n-14 n n r bronlck4oih ,.woFtrA Tiprn.tn benoef4 r. b..e V9x_x1i XVz_OV~~~~~~~. W LS L VX_A.. .L fJL LOf L zsVKFS~ VflL xU 10 V 1 l.a resulted from elimination of this type of flooding. Construction forces are already ..obilized (..ainly manual labor recruited locally) and LrWnru- is anxious to complete the 17 polders during the 1970/71 and 1971/72 seasons. 1.I =ethI&e r o-Jrnot "wat=er& managCemMent (se-- par -3.1 I ca be applied rapidly to the Phase I Group A polders is in doubt pending further studuy. "I';le costs iLn TlabUle IL prov'lde Afor (a) com,p"letiLon of' th'e er,i'anr, EflltmnS in these polders (cost $40 million total equivalent) and (b) studies of iInterior water managem.ent (cost ';1.4 ,,1,1ilon --- al e-uivalCuL, out no capital investments for this purpose pending the results of the studies.

3.08 Phase_l, Croup B: These 8 polders, which cover 0.5 million acres, are in zones olf relatively 'fresh water. Since farmers now depend on these fresh water inflows and since there has been opposition to construction of these polders, it seems clear tnat for tneir development "water management" (para 3.1.1) will have to be planned concurrently. Funds programmed for Group B in Table 1 ($40.0 million total equivalent:) - 1.8 - would bJe utilized primarily fLor pol. er-by-polder planning, 'or tLie first two years (170/71 and 1971/72) and for construction which wzould start abouL JanIuary 1973 adlU Lmight L1AULbUe cmP k Lte .n1 197 V/ 7. TIhe actUai LrtU of implementation would depend largely on the resolution of organizationlal and institutional questions (see para 3.13). ln 'able 1 prograi:miing .,as been assumed as follows:

Cost Start Complete (Imili-ions)

Poider by polder planning 1/7/70 30/6/72 3.3

Stage 1 construction 1/9/72 30/6/75 18.3

Stage 2 construction 1/7/75 31/12/78 1.8.4

3.09 Phase II: Considerable parts of the Phase II area are knowrn to be in freshwater zones. It is likely therefore, as in the case of the Barisal region (see Cihaoter 4), that low-lift punrping should constitute the initial stage of development with embankment construction coming later. Table 1 indicates an initial period of project planning during the years 1971 to 1974 ($1.5 million total equivalent) with capital investment starting September 1972 and extending (in two stages) until June 1978. As for the Phase I Group B polders, water management would be concurrent (or might precede) embankment construction. The sum of $60.0 million (total equivalent) is estimated to be enoughi to cover the full capital investment cost for about two-thirds (i.e. 670,000 acres) of thie Phase II area including interior water management as well as flood-prevention embankments.

3.10 Overall management assistance: An overall management assistance program, including all facets of implementation from management of water supply, operation of physical works and agricultural extension througil advisory services at all levels has been programmed for the entire Coastal Embankment area as indicated in Table 1 ($7.7 million total equivalent). The management assistance program, together with the project planning studies described above, would also have as an objective the development of a viable organization to direct the overall activities needed for development of the entire 3 million acres in the Coastal Embankment area.

3.11 Water management: As hias been indicated in para 3.01, `tJater management" and the provisions of agricultural innuts are needed to realize the full potential of the Coastal Embankment area. 1/ Water management

1/ This has been the opinion of several technical groups including the consultants engaged bv EPWAPDA (Leedshill-Deleuw) and the ECAFE Advisorv Group on Development of Deltaic Areas; see their reDort of February 1966 "Appraisal of Some Aspects of the Coastal Embankment Project of East Pakistan'. - 19 -

would ihave several facets including: (a) introduction into the polders of non-saline river water and its distribution by various means; I/ (b) conservation of natural rainfall mainly through adoption bv farmers of proper cultural practices; and (c) development of deep as well as shallow groundwater to the extent available. Preliminary information with respect to the extent of areas that could he developed by the first means indicate that it might cover 1.5 million acres or ahout half the total area of the coastal embankments. Closure of some of the tidal estuaries might extend this area considerably but the feasibility of such schemes can be determined onlv after detailed study (see water study of the Southwest regicn described in Chapter 5). The extent to which the other two methods are an].icable is not known at present but would be determined in the course of the planning studies described in paras 3.07, 3.08 and 3.09 above.

3.12 Or anizational and institutional problems: Although the problem of water management is in part technical. it also has important organizational and institutional aspects since to solve it will require close coonprntion among EP1WAPDA, the Department of Agriculture, the Basic Oemocracies and Local Government (HD & LG) Department, the ADC and the '[hiani Trrication Provram (see Volume T. Chanter VII).

31 Organyii7Otionnl and instit,it-Innon nrnhIpnmq arp im,n1uvle not only with respect to the implementation of water management but also as regards the provision of agricultural inputs. These matters ma he somewhat more complex than elsewlhere in East Pakistan owing to the remoteness of the area and generally poor communications. LIand-tenure factors mav also be a complicating feature as there are many relatively lare oliAngs and share-cropping, is quite preva-lent. The study ie thus far to the organizational problem (mainly by Leedshill-Deleuw) has inLdated, as a possible approach, thIe se1t1-g urp of a series oi0 "0T",orkLin Units" ancl "Task Forces' which would deal both with water management and agricultur.al 'nLputs. S..ce dloes not perr,,lt a full. UescriLpt0Lon ofL th.Le scheme, which still needs further discussion and acceptance in any case. A4 suimuTnarv of th"e sch"eme is as follows:

a. A "7orking Unit" would deal with 2000 acres at a time and would consist of a Planning Unit, a Construction UInitt an-' an an r M IL L0ILnILI ULI;LL oLULaJ. peLrUsIoIn.l J32 o which, 3 professional level, 8 sub-professional and the rest cLeriLcal anlu/lor unskilled).

1-/ By manipulation of gates on drainage sluices, fresh water can be allowed to enter most polders during high tide. This is however not as easy as it sounds. There are many variations and combina- tions of factors involving characteristics of tlhe tidal cycle, amount of fresh water discharge, degree of salinity and local topography. Thus each polder will require detailed study. -- 20 -

b. A "Task Force" would supervise the work of 5 Working Units. The Task Force would be comDosed of 7 officers as follows:

1 Leader 1 Assistant Engineer 1Thana Arriculture Offircr 1 Thana Irrigation Officer 1 hanna Cooperat-iu Offirpr 1 Circle Officer (Development) 1 RenrpQpntntiup nf Pakistan Academn for Rural Development (Comilla)

c. The 5 Working Units under the Task Force would deal withi only 1if)000 acrrps initailly but each year they would move on to new areas and eventually (during a period of perhaps 10 to- I1 yan-ear woiiuld rcover 1Vn'00( acres

d. There would be 5 Task Forces and 25 Working Units initially. This would be expanded gradually to 20 Task Forces and 100 Working Unlts soz fas, even7r eve,pnt-iua11xy Hi- entivre 3.0 million acres in the coastal embankment area. The total manpower requ4red would t hus be:

I14 0Officers (for 'O 'T'ask-Forces) 300 Professionals (for 100 Working Units) 800 C.0SubD-professional's (for 100 WI.-.orking Units) 2,100 Clerks and unskilled (for 100 Working Units)

e. The initial 5 Task Forces and 25 Working Units wouLUd emiploy r".Lot PolUders" as a trlinlng and .UIlloU0nsL[1U1 device. One such pilot polder is already in existence.

3.14 Relationship with Southwest Region W4ater Study: The proposed Southwest RegLon Water Study would be carretdU ouL conUcLurreLLy wirLL[tLthC studies and investments described in the preceding paragraphs. Since, however, the water study would take about rive vears it wouid not greatly affect these studies and investments which would be well underway by thie time the water studv is completed. On the other hand, the studies and investments would produce a feedback of information that would be utilized in the water study.

3.15 Special Bank Mission: Work in the Coastal Embankment area thus far has been carried out almost exclusively by EPWAPDA and its consultant, Leedshill-Deleuw. From the foregoing discussion of organizational and institutional problems, it is clear that solution of these problems will require close and intensive inter-agency cooperation as well as continued use of expatriate consultants. The precise methods for achieving the - 21 -

needed inter-agency cooperation has however not been worked out. To assist in formulation of a possible framework for such cooperation, the Bank has offered to send to East Pakistan a special team that would be composed approximately as follows: an agriculturist, an agricultural economist, an agricultural engineer with knowledge of irrigation and drainage, a rural sociologist and a public administration specialist. The tasks of the tearn would be: (a) to conduct a reconnaissance in the field; (b) to conduct discussions with the various GOEP agencies concerned and to obtain their preliminary agreement with respect to the organizational and inst:L- tutional setup that might ultimately be desired for the entire Coastal Embankment area; and (c) to draw up terms of reference for the consultants that would be needed to assist in the project planning, actual implementation arid overall management as has been discussed above in paras 3.07- 3.12.

3.16 The above three tasks would require a time period of about a month. Another month would be needed for preparation by the team (in cooperationi with the East Pakistan Bank Group) of a report setting forth conclusi o!1S and recommendations resgarding the organizational and institutional setup, and the terms of reference for the consultants (items (b) and (c) in the precedino paragraph). In view of the nrevious discussions that the team will have had in East Pakistan, it should be possible to reach final agreement reaarding the concu-usions and recommendations of the team's report in a fairly short time -- say, two months. Steps required for the en-a-ement of the consultants as well as arrantements for their financing could proceed concurrently and it might be possible for the consultants to be-in their work by the beginning of 1971 or shortlv thereafter (a starting date of January 1, 1971 has been indicated in Table 1). As described above (paras 3.08 and 3 09), the mainr cnnital investments for Phase I, Group B and for Phase II would not be needed before January 173 4in both cases.

r --- K'obad-a Pro-ect (Kushtia Unit)

3.17 AS -l-ready- mentioned (para 2=09c) the inueStment made to date in the Gl-Kushtia project lhas been substantially under-utilized. Although the technical difficulties that impede efficient operation of the project are graduallv being solved, a stepped-up effort is needed as well in the organizational and management fields as in the technical field. It would then be possible on a progressive basis to achieve a substantially large increase In the area rrigated and thereby the benefits from the project lo accomplish this it would be necessary to employ a consulting firm to be charged withl m.anagem,ent of the project which would include a- detevmli- nation of how the project must be modified so as to improve the operational efficiency of facilities to prevenLt waste and losses of water, improve water deliveries to the farmers, improve use of water on the farm, train personnei and establ ish repayMent for water deliveries on an equiltable basis.l/

1/ Province-wide approaches to the problem of project-cost repayment would be part of the Comprehensive Study described in Volume I. - LL -

3.18 Agriculture in the area is in a transition period. Farmers are very slowly shiLrting from Ltle! old ral.-fed anu rISKy SUDSISLeHnCe arrilng to a more modern irrigation farming. Irrigation, drainage and flood nrevention offer the possibility of replacing the old monsoon pattern of broadcast mixed aus and aman paddy on intermediate land -- and broadcast aman on low land -- by broadcast or transplanted aus, followed by transplanted aman. The following paragraplhs contain further detail on the current problems concerning the GK-Kushti.a project.

3.19 Phase I: The construction of Phase I was started in 1955 and the physical works are substantiallv completed. The essential components of Phase I (net irrigable area 120,000 acres) are:

(a) A pumping plant near Bheramara, consisting of a main and subsidiary pumping station together with a power plant.

(b) A canal. system consisting of an intake channel, a main canal with appurtenant structures, and a system of secondary and tertiary canals and field distribution channels. At about 6.5 miles from the pumping station the main canal -- called the Ganges canal witlh a capacitv of 5,400 cusecs -- bifurcates into a branch canal called the Kushtia canal (capacity 1,200 cusecs) which conveys water into the Phase I area. The other branch (4,600 cusecs capacitv) is still called the Ganges canal initiallv and, further downstream, the Alamdanga canal; It conveys water to the Phase II area.

(c) A flood embankment along the Ganges and Gorai rivers protecting both Phases I and II from inundation by the Ganges river.

3.20 The major problems requiring solution with respect to Phase I are summarized as follows:

(a) Up till now the number of acres actually irrigated is around 40,000 acres of which the bulk consists of transplanted aman. Thie area irrigated during the dry season -- the winter months December, January and February and the summer months March, April and 'ay -- remains very small (only about 4,000 acres). The main reason is that the farmers have no confidence regarding the reliability of the water supply and choose to grow rabi crops (mainly pulses and wheat) on residual soil moisture. During the early kharif season (April-June) the most suitable crop would be transplanted aus of an improved variety, such as IRRI aus from which a yield of over 60 maunds could be obtained. This is, however, a delicate plant and only the larger, more progressive farmers would be willing to risk planting it. The next most suitable crop would be a local variety of transplant:ed aus giving a yield of 35 maunds. However, the farmers, in general, prefer to stay with rice crops basically dependent on monsoon rains with only supplemental irrigation from the project. They are thus not prepared to plant rice crops, such as transplanted aus, that would be largely dependeni: on irrigation, at least during the pre-monsoon months of April and 'May. As already stated, this is because they do not as yet have confidence regarding the reliability of the water supply.

(b) The reasons for the unreliability of the water supply provided bv the Ganges-Kobadak canal system are: first, the main pumping station is still not functioning proper- lv and. second, the canal system is not able to operate at partial capacity. The current problems concerning the pumping station may be summarized as follows:

(i! TWo Dumphouses, the main DumDhouse and the subsidiary pumphouse serve the Kushtia Unit. The main ntumnhouse- comnrislncg 3 Hlitachi turbine pumps each of 1,300 cusec capacity has a 10-vpar historv nf difficultiesp Pump Nn 2 was commissioned in May 1969 and Pumps Nos. 1 and 3 in March and Anril 1970 rpnnprtivelyv Structural settlement problems affecting the t-hree pnumn uinitst'rmnian unrPnsluPdL Tnst-alle-. tion of some auxiliary equipment is till out- standinng.

(ii) Theh cihcidliary, niumnhni,,zp e-mnric - 12 Rtnrk mi yed flow vertical pumps each of 116.5 cusec capaciity. This installat-inn was: orioinally rnct-rite-te4 !:o provide temporary water to the developed sectors of Phase I on an interim basis whilst awaiting completion of the main pumphouse. It now suffers 4 fro-m frequent- hvealcdnym. requi,, res const ant-nI In. tenance and is considered unreliable.

(iii) The power supply for the pumping station remains shaky. A report prepared October 1969O by DWADIIA' general consultants suggested three alternative solutions, which one involves ob-ain-ing polwer ~ I~, , of LtL- tJLU ~ tUL±UL*~f~NA.VJ from a 90-mile long 132-KV transmission line which is iiowever considered an unreliable.solution. Tihe other two alternatives involve installation of an aduUi tLional genLerator at thie DiieraiIiara pUwer station which is adjacent to the pumping station. - 24. -

However, this may also be less than acceptable considering that the station is rather old and does not appear to be well maintained. One of the two existing gnnprntnrc has been out of operation for two years. If the Brahmaputra River interconnector were constructed the nroblem of power supply might be easier to solve; however, the status of this interconnector is now in doubt and in any case several vears of construction time would be required. (v) Trhe nlntzke ch-annel f7rom --he Ganges R'iver, servinEg LV I aI itLrc&_ aihI £ I. . U I L-tkfflrt ttVL both pumphouses, is subject to heavv siltation andU mIustLbUJe dUreduged lUetween OLctober anad Marc'. annually. Annual deposits range from 12 to 18 f eet lIn thickness, so tLhat water supplv to Lhe Kushtia Unit must be regarded as unreliable during the dry winter season due to intake channel blockage by siltation. The dredging of the channe'L consitutes moreover a heavv annual operating cost and spoil disposal areas utilize valuable stud'y of this probleml is -needed .LL. land,.* A11l specialb -UL bL ilZ i1_ IL pt £I. t Ii itta.. including a physical model of the intake clhannel to ULeterLI-ILne hIow best Lto redAuce the existiCngL serious siltation.

(c) The project canals (which are intended to distribute water to the project area ent'rely by gravity) have been desi-ned in such a way that full command can only be obtained at full supply level. Basic to the canal design has been the introduction of "downstream water control" to be effected bv self regulating hydraulic gates anu lower category canal offtakes directly downstream of these gates. For a number of reasons this system has not been successful. Thlerefore, in periods of low demand, the project cannot be operated with partial supply, but has to be operated on a Dart time Dasis. Th'isi is considered unsatisfactory bv the water users and moreover results in high operational losses. To run the project with partial supply would require raising of the canal banks. It would appear urgent to do this as otherwise it would be difficult to make a start in providing a reliable water supply especially during the summer months.

(d) Seepage losses in the major canal system nave been very large. The situation has improved in the course of the years but accurate data on the canal seepage is not available. - 25 -

(e) The detailed irrigation canal network shows certain deficien- cies, in particular with respect to the capacity of the tertiary canal offtake-structures (outlets to the fieldchannels) whichi under development of a modern cropping pattern appear to be too small (see also (g) below). These have a capac:ity of only one cusec per 100 acres which is much less than what has been found (for example) to be needed for boro irrigation with low-lift pumps, namely 2 cusecs for 50 acres. Also the number of these structures may have to be increased t:o nromote an efficient distribution of water amongst the farmers of an outlet unit.

(f) An assessing board, working in the years 1966 through 19458, estimated that the operation and maintenance costs of the project would be from Rs 70 to 80 per acre. The board recommended a water charge of Rs 3 to 21 per acre. No charges have been collected so far. The collection of water charges would seem to be important not only to provide revenue but also to provide an economic incentive tending to reduce water wastage and improve irrigation practices.

(g) Recent studies of cropping patterns and water requirements indicated that no more than about 60 acres can be irrigated during the "summer months" of April-June (also called "early kharif season") per cusec of main-canal capacity.l/ Since the capacitv of the Phiase I canal is only 1200 cusecs, this would limit irrigation during these months to about 72,000 acres. The canal capacity of 1,200 cusecs would however be ample to irrigate the entire Phase I area (120,000 acres net) at other times. This raises the question of whether tubewells might uiltimately be a valuable supplement to achieve ftill summer irrigation. Tubewells might also provide an economic means of achieving re-use of water from the main pumping station which is now wasted either from canal seepage or from farm losses due to the flood-irrigation methods now utilized. Further detailed studies of soils, cropping patterns and water requirements are needed before

1!lhe number of acres irrigable per cusec of lateral-canal capacity is actually unknown and may be substantially less. This is a matter rprinirinv further study. Farmers in the GK-Kushtia area. as elsewhere in East Pakistan, practice "flood irrigation" whereby indivi- dual nlots are flooded and downstream nlots often eet no water until the water demand of upstream plots have been fully met. This kind of flood irriantinn could theoretically be eliminated by installation of a rational system of field channels but in practice this is rarely attainable owina to the small fr-cmPentd Iandholdings (not as hbd however in CK-Kushtia as elsewhere in the Province): see also para. 2) 14). this question can be answered; it might however be practicable to obtain interim answers indicating that there should be no delay in installing tubewells in those parts of the project area that cannot be served by the canal system.

3.21 Phase II: A recent review of soils information has indicated that (a) some clearly defined areas should be excluded from irrigation e.g. the area designated as Bhairab association with a poor irrigation capability; and (b) some overall reduction of the remaining gross area should be allowed for, in particular, ridge-land association and small un- economic water management areas. This reduces the net area to be irrigated for Phase II from 230,000 to 195,000 acres. If Phase I takes 1,200 cusecs and if the nrimarv pumDing station canacitv is 3-900 cusecs (subsidiary station for standby use only) the capacity available for Phiase 11 is 2-700 cusecs. Using 60 acres/cusec for earlv kharif irricarion as for Phase I, gives 162,000 acres irrigable in Phase II. (The balance of 33,000 acres could presumahlv hp irrigatrd hv tuhpewells=_) Hnowever, as indirated above, the primary pumping station capacity is not yet firm. The extent nf ePnwnace in the Canes: and the Alamdano:a canals (nara- 3.19) as wsell na the other Phase II canals comprise further unknowns. The resolution of these problems as well as nrogress in organizing farTnerst outlet aromms may take some years to work out. A possible way to proceed could be to aclhieve at least partiail ue nf the evictina farilities hu aimitt1ng water into the canal system even if only to ground level and then lift water out of the canals b- portable low-lift- pu-p.

AS Asof February 1970, the ststus of constrtinfor tepsc works of Phase II was as follows:

Main Canal (Start 1960-61) 75 miles 7O% Secondary Canal's ({Start 1960=61)8A r .. miles 40% Tertiary Canals (Start 1961-62) 272 miles 14X Fi eld "'arnre's 160(I:N'os. 0%I Drainage Channels (Start 1961-62) 457 miles 62% HlyUdrul1Uc ;,tructures (Start l9766-6) Regulators 203 Nos. 36% Escapes 9 ,30s 59% Syphons 236 Nos. 129

3.23 The wide dispersal of effort over the entire Phase II area is noteworthy. On orders from the Chairman the number of Divisions assigned to complete the work was doubled (to 6) but commensurate funds to utilize this increased manpower capability were not made available. Consequently personnel have been substantially under-employed. - 27 -

3.24 Agricultural extension: EPWAPDA now provides agricultural extension services in the CK-Kushtia area. EPWAPDA's extension service organization is headed by an Extension. Director who is assisted by two Extension Officers. In addition there are two Agricultural Officers, and two Extension Overseers in each Thana and one Extension Overseer per 1,500 acres. Training of model farmers and managers of cooperative societies has been initiated. FAO established at Baradi in 1967 a training center for the purpose of training lower grade extension workers.

3.25 'lhe change from traditional crop patterns is not being effected easily anid it is acknowledged that education of the farmer is a pre- requisite to success of the project. Cooperative societies have been given assistance covering all aspects of farming such as credit, irrigation and drainage. Cooperative societies, the organization of which has been started in a number of Thanas meet weekly and are guided by the Extension Overseers. 'I'hese Overseers constitute the link between the farmer and the EPWAPI)A operating and maintenance staff in the field. While recogni- zing the essential nature of agricultural extension to success of the project, it must also be recognized that until and unless a reliable water supplv can be provided during the rabi and early kharif seasons, efforts in agricultural extension alone are likelv to be considerably less than would otherwise be the case.

3.26 Conclusions on Phases I and Il: These may be summarized as follows:

(a) It is important that benefits from the Kushtia Phase I Unit be substantiallv increased at an early date. It is likewise important to get at least a partial utilization of the Kushtla Phase TI Unit.

(h) 'Terms of refprenrp shoulnd he nrpnarpd for the engagem.ent of a consulting firm covering the fields of engineering,, agricutture- aorieultural economics and nroiect management. The firm would be charged with the studying of the prnecrt and exneditine means of oett-nc the desired results. In preparing tlhese, the terms of reference in the Bank's November 1967 mission report as well as the comments made herein should be taken into consideration.

(c) Collection of water charges would promote efficient use of the prolect while providing needed revenue. See para. 3.20(f).

(d) The use of groundwater, if only on a supplemental basis, should be conidered 4,' hoth th_l KuShtia Phoose Tao -A Phase II Units; see para. 3. 2 0(g) above. Use of ground- wat- .n any part of thAe (CKarea should not be construed as indicating a lesser need for diversion of surface -- 28 -

water from the Ganges River. This matter will have to be studied on a broad basis to include, for examnle, saline- intrusion effects. Such a broad approach would he followed in the proposed Southwest Region lhater Studv; see Chapter 5.

(e) The costs indicated in Table 1 are for the following:

(i) Part (a) involving $6.0 million total equivalent would cover about 1,000 low-lift pumps supplied from the existing canal system and 200 tubewells plus some improvements of thie existing canal system to enable drv- season irrigation of about 50,000 acres within a few years.

(ii) Management assistance including. engineering2 and agricultural aspects as outlined above; cost ;2.0 million total equivalent.

(iii) Completion of Phase II at an estimated cost of bt6 million total equivalent. This work would be scileduled to start July 1. 1973.

ChandDur Proiect

3.27 Some description of this nroiect lhas alreadv been given in paras. 2.13, 2.14 and 2.17. The major works of tlue project include a nperipheral embankment 60 miles long; from 10 to 15 feet high anid Aitlh a crest width of 14 feet (wide enough for vehicular traffic); an irrigation cum drainage p.umping, station with 6 uni ts having aArnmhind cnacitv Cof 1,135 cusecs; an outlet regulator (for drainage) containing four 30-feet wide by 15 feet high gates; and 2 navi,atioln locks, one adracent to rIhie pumping station and one to the outlet regulator, each 100 feet lonxw and 20 feet wide.

3.2t .As already mentioned (para. 2.13), irrig-atlon distrib::tion will be accomplished by making use of the natural drainage channels. Wlhen exLterLor water levels are suff-ic4lently high, hLese channels w4llLe filled by gravity with water admitted by operating gates at the pumping station as wel'l as thle outlet regulatuor ,ates. ,1 e e r leve ar l the pumps at the pumping station will be operated. The water will be LiteLU LLroILm thLe drainage channels vbyUdeseJ l powered- 2=cisec pumn.,s operated by TIP irrigation groups (see para. 2.02).

3.29 Tllirty-one miles of the peripheral embankment have been completed to date. Contracts were awarueud Ln early 1970 f.Uo Lthie pumpiLng station andU outlet regulator and for the mechanical and electrical equipment involved; these structures govern the overall construction schedule and are exxpected to be completed bv the end of 1972. - 29 -

3.30 Tlie planning, design and organizational conceDts that have now been embodied in the Chandpur proiect (paras. 2.13 and 2.17) ar expecte-d to find quick acceptance by farmers of the benefits resulting from comprehensive xwater control (irrigation, drainage and flood nrevention). A principal reason is that with water actually available throughout the proi,ect area ('via the drainaoe suystem), farmt.ers will actually see waster. during the dry season; this should convince them rapidly regarding the reliability of the pro,ect 's water supply -- a feature that has been lacking, for example in the Ganges-Kobadak project described above. (That farmers react no s i tivel- to read- availability of ws.tertduring the .J- season hias in fact already been proved by the success of the low-lift pump program; see para. 2.01.) A second reason is that the use of the Arainage channels means that far less land acquisition will be required than for a svsllemo wravity canals; this shoUld also pror.ote fa-rters' confidence in this zone of extremelv dense population. Finally, the fact that 5.JA it ;wil not be eli)4,inateA loo abruptlYv will gve 'ariiers ;'me to readjust their kharif season cropping from broadcast to higher-yielding transp'lantedl amEan.

3.31 UaIdnpur 'as been able to ga'n in other ways from the Ganges-- Kobadalc experience. The pumping units in the Chandpur pumping station wi 11 be of relatively srwpiLe uesign -- perhaps a Dit less efficient in a strictly technical sense than those of the Ganges-Kobadak project but on the other iaUid easier to maintain and operate and better suited to East lakistan conditions. The Chandpur pumping station will also be located. on a relatively stable water channel (the Dakatia River) and will not suf-fer the serious siltation problems that have afflicted the intake of the Ganges-Kobadak pumping station.

3.:s2 The appraisal report, on the basis of which IDA in April 1970 granted a credit of US$13.0 million and which is included in Volume IV, indicated that the economic rate of return of the project would be about 22 percent. In this calculation the cost of farm family labor was included; if omitted the rate of return would be considerably higher. The anpraisal report also estimated that the net value of production per acre would increase from Rs 425 without tihe project to ks 1,411 with the project, an increase of 232 percent. Despite the increase and considerinp that the average farm holding is only 2 acres per farm familv, the facr remains tihat even with the project the standard of living of an average farm family will remain low.

3.33 With respect to recovery of project cost, under existing TIP procedures, the cost of operation, maintenance and replacement of low-Lift pumns are to be fully recovered by a 'rental charge" collected by the Thana associati-ons. In agreeing to the IDA credit, the Government assured IDA that rental charges would be collected on a graduated scale rising to a level sufficient to ensure recovery by the tenth year of full annual operation, maintenance and replacement cost which will amount to Rs 47 per - 30 - acre. The cost of fuel estimated at about tihe same amount per acre will be paid di4retly by tl:e pump group members. l',e cost of operafioilan' maintenance of tile major project works is estimated at about ivs 13 per acre; tl.e G-vernmer.t agreed t1-t- tLh- y -- A dtermineA- W.t i-. a; l=mon.'i period how this cliarge could be collected.

3.34 With respect to repaymeat of capital cost, tLle Governmenit agreed Lo give th4s further study. In actuality tiq i l .or par. of a Province-wide investigation to be assisted in part by the Compreiienlsive

de..cr.0. ,_A VLol:ume I.

3.3 Tr.;iefcact '"a' Ltie Char.dpur project area i led esl ou -, *.JJ LAA LUL -LLLLLA U.,I OL'AjUL JLL aLj=% LZb 1.LCLdUY UCLL23t.LY P)~UPU lated and that tihe population will continue to rise present a problem whicn is not limited to Llie "hlandpur area andiu is Lyp-ical ovi Las. Paklsta1 generally. There will be contiLnued pressure on the land not only for agricultural production but also Lor additional rural h OUSinLg and for roads. These problems wlhiclh lhave no easy solution will require furtlher investigation and long-range planning as described in Volume i in the chiapLer on the Comprelhensive Study.

3.36 The planning concepts involved in the Giiandpur project represent nevertheless an important step forward. It shiould be borne in mind tilat a completed comprehlensive project suclh as Cliandpur does not yet exist in Last Pakistan. Successful and rapid implementation of tLe Chandpur project thus lhas crucial importance since it will assist in bringing to ligiht problems still requiring solution with respect to the many projects simiiar to Clhandpur to be undertaken in the near future. CHAPTER 4

NEW PROJEGTS -- SHORT RANGE

Low-lift Pumps

4.01 The history of small-scale irrigation including low-lift pumps has been described in Chapter 2. To assist in the evaluation of the potential for low-lift pumps, the East Pakistan Bank Group in 1968/69 maldef arn eval7.uat4.on of the quantity of surface nwater ava-ilable for suc.l. pumping. Two criteria were used: (a) a mileage criterion whereby 5 low- lift- pum-ps wouldI be placed per. mille of su4table and4 accessible streau,ba,,k (many portions of the streambanks, especially of the principal rivers, are not' for loatn lowlif.- pur.p b4_ecause I5IJUsuitabl OU. LO ±~ LIL LU%La LA.L16 WWJW .L&. P UF. II I.J.ILof sand1OIU accretions;U L . A~L.ULAM L.LitICt-htse often have transverse dimensions of from several hundred feet to more ti-an a amile, ~~ WU±I.l~which Ibis_ beonUr-YLJI%LAIC th opralnV0IMLCLLA.11 LLCFtU.LL.±LLYcapbiit -- of4IVI the I LL I---lfUW-iLX12=8san L pumps. iiiiu (a) a flow criterion whereby no more than 60 percent of the dry-weather Low of1 tihe streari-i would bUe utilized at aniy one t'ime with thne otner 40 percent reserved for other purposes, namely fishing, navigation and salinity control. The mileage criterion was tne most important as it covered 82 percent of the cases. The estimate indicated that the placement of 32,000 low- lift pumps could be justified each with a capacity oI 2 cusecs.

4.02 Typically, the low-lift pumps are used for winter irrigation of boro rice. The zones where perennial winter water is available are also generally zones of flooding and it is necessary therefore to remove the pumps before the monsoon season starts. The extent of supplemental irr:L- gation for aus and aman rice is thus rather limited. If it were practicable to operate the low-lift pumps 24 hours a day during the months of peak irrigation demand (January and February) and to shift their location, and provided that water could be distributed with reasonable efficiency so as to minimize wastage, it woulu theoretically be possiDie to irrigate perhaps as much as 100 acres of boro rice with a 2-cusec pump. In practice however, because of the small fragmented landholdings and the practical difficulties in shifting pump locations it has not been found possible to irrigate more than 50 acres with a 2-cusec pump and the current average is actually about 40 acres.

With a spacing of 5 low-lift pumps per mile, the width of the irrigated strip paralleling the streambank under the mileage criterion is about 4,000 feet. It would appear that in many cases, depending on terrain characteristics, the width of the irrigates strip could be doubled provided the pumps are operated 16 rather than 8 hours which would indicate that even irrigated per pump. However to do this a way would have to be found to cross the 50-acre area of the original pump group so as to make water available to the land of a second 50-acre pump group which is 4,000 feet away from the water source. Considering the right- of-way difficulties involved in earth canals and the high cost - 32 -

and technical difficulties involved with concrete distance of 4,000 feet). it might be possible to develop instead a scheme 7 to 8 feet would be needed. The flumes would have to be carefully staked out in the field and various special problems such as crossings would arise. To handle thestechnical matters, it would be necessary to create some sort of a bureau of agricultural engineering under TIP. Rural roads have been constructed in recent years by the Rural Works Program and it might be possible to combine the construction of such brick flumes with ruiral roads with the construrtion of both carried otit by the Rural Works Prooram. The Bank's East Pakistan Group intends to explore these matters further 2s it might be nnpqihle at relatively low cost to achieve a cnsIder- ably larger area of boro rice cultivation; see para 4.04.

4.03 As has been explained in Chapter 2 (para. 2.14) the use of low- lift pumps as part of larop-scalpechempes (executpd by FPWAPPA) now cons- titutes a principal feature of these schemes which are described in detail later in this chapter. The aic cnc etnnrnis to provide crruc'tu,res (ca;l lesd "regulators") or pumping stations to supply water to the natural drainage suytepm fronm which it is theit distributed to farmers' field channels by thle low-lift pumps. The large-scale schemes are within specific geographic areas and the nuestion arises whether or not the use of low-lift nurms can be extended elsewhere by means of such structures which need not always he of great size. For examlpl in thp large tidal areas of the Southwest Re- gion (and to a much lesser extent in the Southeast Region) it is possible to close many natural channels br nmans of a small harrier (earthen nplug or dam) equipped with a sluice gate. Water can be admitted during high tide and the gate then closed to retain water (as the tide eve1 falls) thus facilitating low-lift pumping. Medium-size structures of this tvpe have already been constructed by EPWAPDA (for example at the mouth of the Little Feni River) but as part of schemes to prevent saline tidal flooding; a valuable by-product has resulted in that stored fresh water has become available for low-lift pumping. The use of minor works to assist low-lift pumup'ng seemis part'Lcular'lv prom,ising ln. Ile loas,a1 E.mhank.ment project area (see Chapter 3) and in the Barisal region, described below. Here, it might be quite possible to proceeid with construction of some minor works along with low-lift pumps as an initial stage, with larger scale works coming later. it has been roughly estimL1ated by AtUDC. that 8,000 additional low-lift pumps could be made effective by the construction of such minor works.

4.04 With 40,000 low-lift pumps in operation, and assumiiing 40 acrdes per pump, 1/ the total area irrigated would amount about 1.5 million acres.

1/ This is the average area served by the 18,000 pumps now in operation. During the coming period of rapid distribution of pumps, the area served ner nump may be expected to remain constant or even drop slightlv and then should rise again if the pump program continues to be well manacgpdr On the other hand. if the scheme described in the indented portion of para 4.02 proved to be practicable, the area irrigated per plm.n rcouild he inrreased tin to nerhans 80 acres. - 33 -

It shotldi be reasonable to attain such objective by the year 1974-75. Thereafter since the area served by the low-lift pumps is as already mentioned mostly in the flood zone, it is estimated that of these 1.5 million acres 1.0 million will gradually be encomnassed within the areas of the EPWAPDA multi-purpose projects.

4.05 Low-lift pumps are supplied by the Agricultural Department Cor- poration (ADC). For each pump, the Union councils as nart of the Thana Irrigation Program (TIP) 1/ organize farners into a group of users before the pumps are supplied. Typicall- there are 20-25 farmers in each pumn group thoughi the number can be as high as 50. As indicated above, pumps are located along perennial streams with suit-ab1le streambaanks, However in many cases natural drainage channels tributary to the streams can also be made use of. Under the TIP, members of the pump groups assist in mapping the drainage channels and to determine where, by means of minor excavation, the-y can be deepened so as to extend the availability of water. The excavation is performed by the local farmers with payment for their labor by the Rural Works Program (see Volume 1, Chapter 7). The fmrvm,ra musQt sorme- times give up some land because of the channel deepening but this is provu'eu at no cost to thLe Govern.ent.

4.016 }1iel channe's 'or water dilstrib1utilon within the typically S!)-acre area of a pumping group are arranged and paid for in a similar manner.

4.07 Besidles supplying the pumps, ADC also operates and maintains them. The pump groups pav a lump sum fLor rental of tLeL purlps to ADC' an' also pay for the fuel. The rental increases each year from Rs 150 per pump in the first vear to Rs 2,700 in the seventh year and thereafter.

4.Wt In 1966-67 and again in 1967-68 the Thana irrigation Progran was carried out in various parts of the Province on a pilot basis. In 1968-69 it was first implemented on a Province-wide scale and was quite successful. By 1969, 11,181 low-lift pumps were put into the field for operation. Due to the late arrival of some of these from aDroad, not all were actually used for irrigation. However, 10,852 pumps with a total capacity of 20,606 cusecs were operated, and they irrigated 424,798 acres. As a result of this expanded area under irrigation, horo rice production increased from 1,113,000 tons of cleaned rice in 1967-68 to 1,612,000 tons in 1968-693, an increase of 499,000 tons or 44.8 percent over the previous boro season.

4.09 The way in whicih the TIP low-lift pump program is organized is important because for the first time farmers are siharing in the cost of irrigationi in East Pakistan. In 1968-69, Rs 1,574,100 were collected for pump rentals. In addition, pump groups purchased 2,771,543 gallons of diesel fuel at Rs 1.38 per gallon for a total of Rs 3,833,009 for fuel. Of tlhis, 55 percent (Rs 2,108,154) represented duties and taxes paid to the Gcvernment

1/ See Volume I, Chapter 7 for a detailed description. - 34 - indirectly. Farmers thus paid a total of Rs 6 million for the Program (excluding their contribution in constructing all the fieldl channels) while the Government's expenditure was Rs 60,000,000. The fact thiat farmers bore 10 percent of the capital costs in the full flrst year of the P'rogranm is of importance because it is the first time thiat East Pakistan's farmers have made any direct financial contribution to a water development project. It also indicates that a system has been developed under which Government subsidies can gradually be reduced as farmers' inconmes and repayment capacity rises.

4.10 There is no doubt that where surface water is readily available low-lift pumps provide the cheapest and most economical form of irrigation in East Pakistan. The cost of a low-lift pumping uniit including pump, engine, accessories and spares is currently estimated at about Rs 8,00() of which about three-fourths is in foreign exchange (assuming the official rate of exchange). Using a figure of 50 acres per lowv-lift pump would indicate a cost per acre of Rs 160. In comparison thie tubewell cost per acre is roughly Rs 1,350 (tubewell and appurtenances only with no allowance for drainage -- see discussion below of ADC 3000-tubewell project) wereas the multi-purpose projects may cost from Rs 1,500 to 1,800 per acre. It must of course be realized that these figures are not strictly comDarable since they do not take into account: the true value andi percentage of foreign exchange (much higher for tubewell projects); the true economic cost of construction labor; and the fact that thie multi-purpose projects provide drainage, flood, navigation and possibly also fisheries benefits in addition to irrigation benefits. Low-lift pumping mav moreover entail additional costs such as deepening of drainage channels and excavation of field channels. In the case of the Chandpur multi-purpose project where internal irrigation distribution will be entirely by low- lift Dumps. the associated costs for such excavations, for engineering, for overhead and for contingencies indicated that the total cost per acre would be about Rs 250 (see anDraisal report in Vol. IV). It is nevertheless still clear that provided additional structures are not needed as dpcrrihpd in nara- 40n irrigation bv low-lift ntimning is the most economical method. A preliminary analysis made by the East Pakistan Bank Group in 1968 (Appendix A.2 of the repnor of Maqrch 1, 1968) indirated a nrobahbl very high rate of return as even if initial costs are discounted at a 50 percent interest rate, they can hb recoeirod in I yanrs. The reonrt therefore concluded that low-lift pumping should have high priority in the

East Pakistan progra. but that wter availability and implementation capacity could limit the rate at which low-lift pumps can be installed.

4.11 In view of the large spread between low-lift pumps and the alter- nativeLVC Ll1_hL11UUb Uf iLLrCLLgai aU1 pFrovijJnd de that the cost of t.he mlinor works is not too high, it appears that the additional 8000 low-lift pumps mentioned in 4.03 are economically Justified.

4.i2 Throughl tihe 19U6/0u wiLnter season, AD'. Chead placedU Ca LLCil Uo 18,000 low-lift pumps throughout the Province. The ADC program has provided for an additional 9,000 pumps to be placea and in operation during 1970/71. The pumps to meet this program plus about 1,000 as spares and replacements, - 35 - have been purchased and will be ready for distribution starting in October 1970. ADC then plans for 5,000 pumps to be fielded in 1971/72 and aboui: 8,000 to be fielded in 1972/73 making an addition of 13,000 and a grand total of 40,000.

4.13 ADC requires financing for the additional 13,000 low-lift pumps. The first group of 5,000 would cost $14.0 million of which $7.0 million in foreign exchange. As indicated in Table 6.1, funds would be needed IDY Januarv 1, 1971. There is a manufacturing capacity for both pumps and engines in East Pakistan. but this capacitv is not adequate to supply all of the engines. The local manufacturing capacity, according to ADC, will be adequate to provide engines for the second group of 8,000 pumps, provided that materials can be imported. In addition, the plans provide for the importation of a small number of larger capacity pumps to be used to feed the 2-cusec pumps. It will also be necessary to construct additional stora,e and repair facilities, strengthen the manaaement of the low-lift pump section of ADC and train service and maintenance personnel. The cost of this second prolect is estimated at USS25 milllon, of which UJS$5 million would be foreign exchange. Since the final increment of 8,000 pumps may rtun into some difficulties regarding water avallahilitv- the fi4gures shnwn in Table 1 indicate a somewhat slower rate of disbursement for the project than that imnlied above.

Tubewells

4.14 Aside from the drv-season flows of perennial streams usable by small low-lift pumps, the large dry-season flows in the major rivers would appear primn farie to nrouide thA most economical source of water for irrigation. A major surface-water project (Canges-Kobadak) was started 15 years ago but for a variet- of reasons as described a-ove has notlproviAed benefts to the extent expected. Feasibility studies for various multi-purpose pro- Jects incorporating irrigation from surface water have been-preare (see descriptions later in this chapter) but detailed reviews of these studies have reveaed di-fficultl- s in im, .entationt- t - :ih nl now are- bt at4aIy being- solved. Since results from tubewell irrigation are obtainable more quickly (but not necessarily more economically) than from surface-water projects, and in view of the urgent need to expand irrigation to realize the full potential of the new Rl'RT zice varieties durin,,g Ile loro and aus season6 , the East Pakistan Government has given the highest priority to tubewell developm-ent inU Ut,cz~~UtIthwe). t Fourth Lt VIlan.tfl- periodt .t-- at*--an tctIhas p.proose J LIl~thIC A10Ln_i ± OL VILiro 20,000 tubewells. Concurrent tubewell programs are underway by both EPWAPDA and ADC.

4.15 There is little doubt tat East Pakistan hlas abundant groundwate r resources whose development is barelv getting started. At this stage how- e e tVtS informtl/lltUfiV~ ~ t. WI; n swAi a0;t sl y,yc'fC e t AtItWt,aA,.i ns u f f 4 .c 1tLC.I-z 1-1 t. LU..L0 plFtLIIJ. _-rm - Lt ±11n tLt11lbiVC e n s L v _- utvtieltpmenLu _ev a- I _u_P _ e a t except in certain areas of limited size. The most systematic overall view "lhus take;; U'S L0t~ttttIt3O LIIfar wasW UItLta-1 9110UC.,ebythe VY LI L'L1ILU".n-ted 0 LdiLttsGooialSFe tO, ot'O igIfCi2d i DU[Vey kU3L~.a) in earlv 1967 (report entitled "Groundwater Resources Investigations Program £orPPakistan", ;as Jul.y 1'7)fj. A0s determined, L[roi thi.s repori an area with a total extent of about 28,000 square miles (51 percent of the Province) conLtailnS Uc-OnsolS IidUaLt U tedaLl tolSuephLDsd of several hundrea feet and. - 36 -

should therefore provide good possibilities for development of large- capacity wells. Outside this area prima facie evi(lence indicates limited possibilities eithier because of the nature of the superficial geologic deposits or (in the coastal belt) becatuse of the proximity of the sea.

4.16 A major part -- 16,000 square miles -- of the 28,000 square mile favorable zone is located in thie part of the Province th,at is normally flooded. Since, in areas subject to normal flooding, surface water generally is readilv available and since pumping will also be required for drainage (thie same pumps can be used in the drv season for irrigation pumping) the use of irrigation tubewells in tthe flooded portion of the favorable groundwater zone appears doubtfujl. The non-flooded portion amounts to about 12,000 square miles but only about five-eights (7500 square miles) is cultivable (three-eighthis are taken by rivers, streams, roads, towns and villages). A breakdown of thle latter figure according to regions is as follows:

Northwest Region 4,400 square miles (2.8 million acres) Southwest Region 2,600 " " (1.7 ' Northeast Region 500 " " (0.3

Total 7,500 square miles (4.8 inillion acres)

4.17 As was pointed out in Chapter 1 withi reference to the Northwest region this is the driest part of the 1'rovince and is also tl,e one wlhere flooding is now considerably less serious than elsewhere, particularly since completion of the Brahmaputra Right Bank Embankment (para. 2.10). Since drought is also more serious in the Northwest Region it is the one where most groundwater development to date has taken place ill the form of the Northern Tubewells proJect (para. 2.12). The ADC 3,000 tubewell project partially financed by IDA and several. EPWAPDA tubewell projects (all described below) will also largely be concentrated in the Northwest region.

4.18 The area of 2,600 square miles in the Southwest region is located in the area of the Ganges-Kobadak development described in para. 1.14. As brought out in Chapter 3 (paras. 3.20 and 3.26), tihe exixting Kushtia Unit of this project, which is based on surface water, will most likely not be able to supply all irrigation requirements and tubewells for supplemental irrigation should tiherefore be considered at an early date.

4=19 In the Thakurgaon intensive-tvDe develoDment (see para. 2.12) over 3 relatively large capacity (3-cusec) wells have been operated successfully ner An,qure mile irrioated. Even if an ultimate fieure onlv half this is used, it would appear that over 11,000 3-cusec (or 17,000 2-cusec) wells nrn1A cafElv hlo lnctnl11d Tt wwould sepm nruepnt thprpfore to nian for such a goal while obtaining confirmation from concurrent investiyations and on- going projects (see para. 4.39) Wlhcn more complete hydrogeaopoic information becomes available, massive tubewell developments (large capacity wells of over 4cusecs and over 500 feet deep) will also ultimately require consideration; see paras. 5.26 and 5.36. - 37 -

4.20 Existing tubewells: The total number of tubewells in East Pakistan at the present time is about 1,360 of which 365, in the Northern Tubewells (Thakurgaon) project of EPWAPDA previously described (para. 2.12), are electrically driven, of relatively large capacity (3-cusec) and closely spaced (about 0.5 mile apart). A more comprehensive development (in the sense that agricultural inputs were provided concurrently) has been that of the Pakistan Academy for Rural Development (PARD) at Comilla where 160 wells have been installed since 1962. Since mid-1966 PARD has also assisted agricultural extension efforts at Thakurgaon. The Comilla wells range in depth from about 150 to over 400 feet, are 6 inches in diameter and have an average capacity of about 1.25 cusecs. Most have been constructed with hand-powered rigs and have utilized locally made brass strainers. The pumps are centrifugal rather than turbine type and are powered by both electric motors and diesel engines. The closest well spacing is about 2,000 feet.

4.21 The Avricultural Develonment Corporation (ADWI tusing itr own forces and local contractors has installed about 800 tubewells in the past two ypars. The cost of t tuihewells to datep whirh have been m.ainly b inclhes in diameter, has ranged from Rs 26,000 to Rs 40,000.

4.22 ADC tubewells project: ADC has just concluded negotiations for t-h-iefinancing of a 3jO0o tubewell pro4ect with a total equvIalent cost' of $44.6 million or Rs 212 million (Rs 71,000 per well). The tubewells woul be incstnl 1 on a turnke-, basis by oontr-actorS (approxmatfely ~~~~~~~~~~~.J 7 - . - . - i - .Wz\M one-third local and two-tlhirds expatriate contractors). The wells will be 8-inch diameter and will be equipped with imported fiberglass screens, casings, deepwells turbine pumps and diesel engines. Sixty percent of the total cost will be financed both by IDA and the SwedIsh International Development agency including parallel financing up to about $7 million biy the Canadi3nIainternational Developm..ent Agency. Both engineerlgt an-, agricultural consultants will be provided under the Credit and the consultants' terms of reference inelude prepara 4J- aAn addiional.^I)C. tubwell project.

4.23 I'he ADC tubewells will be of lower capacity (1.5 to 2 cusecs) than the EPIWAPDA ones described below and they will be dispersed (spaced a minimum of about 1.5 miles apart). Prime movers will be diesel engines. (Som,e of theC ADC tubewlls are and will b-e located in thec same zones as Ile ~~ LII,:. nLj L Ic ,.a L L &l - 1 L . II. LJL U ~LIZ LI a c LUL1U lO LILI EPWVAPDA tubewells but this is considered beneficial in that the ADC wel:Ls provide valuable informa .aion on the aq uifer characterls"ics 'or use -inplan. ning the intensive EPWAPDA wells. On the other hand, it is obvious thal: coordinatior. of the 1r y .) 0prcn.o C,J,IU .L~LA.L IJ L L 11I~L twoI_ program,sL' II .-i.8- necessa tiF)AprxmtlMy*j UAAI _LIfLd=L Y UVJ CLUCI. UL the 3,000 will be located in the Northwest region and the remainder in the nortLiwes; corner o[l tllZ-e INLorthileast region. At a spacing o£ 1., miles, all of the available area (4,900 squire miles -- see para. 4.16) would be utilized. it is contemplated LnaL as uevelopment of tne 3,000 tuDewells proceeds, information will become available on aquifer conditions which wiLl provide a basis for closer spacing of tubewells. (As has been mentioned, in the Thakurgaon area, the average spacing is ahout half a mile; there are indications that in this area welis can now be addedi thus reducing the spacing below 0.5 mile.) - 38 -

4.24 Table I indicates thie Tubewells I (ADC) project as starting December 1, 19701 with completion December 31 1 7e, Th Uttal equivalen cost would be US$44.6 million of wlhich $22.8 million in foreign exclhange. Th±e appraisal repor t containing furt'lier detaOils on this proJect is includ'eLi in Volume IV. Table 1 also shows the Tubewells II (ADC) project starting 2) years later, thilat is, December l, 197' with co,ipletoion Decer,,bler 31, l',76. The estimated total equivalent cost is $84.0 million for 6,000 tubewells. It is expected that many of the tubewells in the second project would be located in the same areas as the first project and would therefore involve a c±Uo[er spdaLigK UI we'1iY. Tb is cuiIs.lu'ereu redsonaDle in View 0f Lhe experience at Thakurgaon. It should be possible, following about one year of uevelopml-ent, to obtain frorm thne first project sufficienIt informationl on aquifer characteristics to enable proceeding with th,e second project; consultants engaged on tnre first project would assist in preparing thle seconld project.

4.25 EPWAPDA tubewell projecc_ts: Four areas are proposed for iarlv development by EPWAPDA. The areas, designated as 'ubewells I in TahJe i, are as follows (see Map 2 for locations): Thakurgaon Extension Tentulia Panchagarh Rangpur North Mrymensinglh 4.26 The first three areas are in the Northwest region while North Mymensingh lies in the northwest corner or the Northeast region. ThtI EPWAPDA tubewell areas in the Northwest have some general characteristics in common; there are limited surface water supplies, monsoon rain tlooding is minimal, and -- based on data of varying reliabilitv -- good aquifers. A feasibility report has been prepared for the Rangpur area, a draft of a feasibility report is available for Tentulia Panchagarli, and a feasibility study Is underway on North Mymensingh. Wells located in the T'hakurgaon area yield data indicating that further development is promising. The principal recharge in Thakurgaon is directly from precipitation; however additional water balance studies are needed.

4.27 The existing feasibility reports contain much valuable information, however, they are inadequate in many respects and will require revision. Additional information is becoming available for existing wells. The ADC tubewell program described above as well as the United States Geological Survey-Assisted program (see below) will also provide information of value. The ADC program raises two points with respect to any EPWAPDA tubewell project to be considered in the near future. First, there should be coordination so that the ADC and EPWAPDA program are not in con- flict. 1/ This means they should either not be in the same general area or where they are it should be recognized that the ADC tubewell program, which has the objective of achieving a rapid increase of agricultural production, is on the basis of a non-intensive exploitation of the ground-

1/ The coordinating agency should be the Plannine Department, GOEP. - 39 -

water resource. 1/ Thus, as a second point, it should be recognized that the intensive development of the groundwater resource in a particular area will ultimately require a much closer spacing of wells and/or large-capacity wells.

4.28 As previously mentioned, the ADC tubewells will not exceed 2 cusecs in capacity. Experience with the low-lift pump program indicates that this is about the maximum that can efficiently be handled by a "pump group". Wells of larger capacity involve some sort of a distribution system (this involves land acquisition problems) and a larger number of farmers within a particular pump group which makes the problem of management more difficult. The tubewells proposed by EPWAPDA would have an average capacity of about 3 cusecs. The average cost per tubewell (includ- inr Pl1ctrir trnnsmission and a 1.800-foot lined dIstribiution channel for each well) would be about Rs 200,000 and the area served per well would be 200 acresr 2/ Thp PstimatPd cnt of an ADC tubpwell inrliiding di4esl en- girne is about Rs 71,000 (para. 4.22). The capacity would be 1.5 to 2.0 cusecs and the area servoed, based on experience with the low-lift nllmn groups, would ultimately rise to 50 acres (for a cropping pattern including boro and aus rice) and perhans 6n acres for one not including much bor- rI4c The lower cost of the ADC tubewells raises some questions regarding the pro- posecl !PWAPDA tubewells, first, regarding economic Jlustification partic- ularlv of electric transmission and, second, regarding optimum capacity of the~ ~~~t well 1-n. wl-41 be in a concentratedA zone anA som.e lin.d distribution canals will be provided, perhaps 3 cusecs is too low a capacity) anrid thLiird thie area se-.^ved per we'll1. 4.2 Thle location, capacity of wells, type of' motor dUrive (electr:Lcal *4. I e lLd L±UA L0 dL L UL W L.LL tp UI 50LU UL V I± .L.L or diesel) agricultural inputs, repayment, 3/ economics, 3/ organization and rranigemeit of1 thle Tubewell I dUevelopiient ali nee furtheL study. ho-wev,sr , based on current data EPWAPDA has tentatively sized the development as

*olOls:

ThlaLkurgaon Lxtenslon 300 we'll 60,I00 a.res Rangpur 300 60,000 "Tentu'lia 510 " 102,000 £ Panchagar'htiL£a4lLIIO,dL L LId IIJ C£Uld4 V i North Mymensingh 300 " 60,000

1,410 wells 282,000 acres

/ A… iJs t.~ Lw_. £ L Aflfl t. 1 1 r ±/ An important aduditonalbeneftWLoL OL tle ADULCtubewells 'is thLat, if properly monitored, they will provide invaluable information for the USGS-assisted study.

2/ From PC-! form prepared by EPWAPDA for 1,000 tubeweii project, May i968.

3/ Needs treatment as for the multi-purpose projects; see paras 4.39 and 4.40. - 40 -

4.30 The Tubewells II project would be an EPWAPDA project comprised of 1.000 tubewells located in an area Drotected bv the existing Brahmaputra Right Embankment project (paras. 1.24 and 2.10). EPWAPDA is negotiating a contract for a feasibility study and final engineering for the proiect with a Yugoslav firm called Ceotecknika. This area shows promise for a good groundwater deuvlnnpmnt hut the feashib41itv study should Include consideration of all the elements and alternatives recommended for study on Tube- wells IT

4.31 EPWAPDA proposed to have the feasibility studies revised for the Tubewell I areas, and have tendering documents and final engineering done by the same consulting firm. Table I indicates this work beginning July 1, 1970; revision of the reports for the first three areas would be com-plete by mid-19171 and for North M--4-sin, by mid-17972 Const-rucion starts are scheduled for all but North Mymensingh by mid-1971. The esti- m,Lated aggregate cos-t of Tubewells I 4 $56-.4- m.llion (totlent based upon preliminary data. The feasibility study and final engineering on Tubewells II are scheduled for completion mim,id-1973 with co.struct'on scheduled to begin at the beginning of 1974 at an estimated cost of $42imilli4on based uponprlmnyda.

43 R.Cate of tubewel_developmer,t: The pri ncipl 1 constraints ot accelerated tubewell development appear to be: (a) The rate of formation of water-user groups (the tubewell equivalent of the low-lift pump groups, both of which are formed under the TIP program) though this might be rapidly overcome with the new emphasis on tubewell development and coupled with the TIP program (see Volume I, Chapter 7);

(b) The speed at which groundwater conditions can be verified in specific areas;

(d) Institutional capacity with respect to supervision, maintenance and proper irrigation distribution.

4.33 Financial resources and trained manpower are not major constraints. Roth thp Center and Provincial Governments as well as several foreign donors have indicated a readiness to make substantial additional funds available for an cP1celratPd tubewell nrogram in East Pakistan. During the construction period technical difficulties will no doubt arise (access to many tubewell sites mayr not be aai lahble during the m nsonnn seaon)* after comnlerion. operation and maintenance problems will grow in scope and complexitv. However, it is 4Aju,diged tohat evicting lorcal ConnrrqtOr capacity; tooerher with expatriate contractors and thie limited but developing capabilities of ADCr, PDW.ADnA a- t-he Co,illa Cnnerat-ive shonuldli b-e sufffiripnt to sink wells as rapidlv as water-user groups can be formed and thie necessary conco ta1rit A hb the - 41 -

water-user grouLps. A confirmation of this judgement should be available by about mid-1972 (tlhe end of the 1971-72 construction season) when considerable progress slhould have been made on the ADC Tubewells I project and, to a lesser extent, on the EPWAPDA Tubewell I developments. The major constraints thus appear to be mainly institutional, i.e. TIP, Department of Agriculture, ADC and EPWAPDA. For the present, it is estimated that the present annual rate at which well groups can be formed and wells sunk would be about 1,000 to 1,500 ADC-type dispersed wells and 500 to 750 EPWAPDA.-type intensive wells. This rate would increase as the key institutions -- TIP, Agriculture, APC and EPWAPDA -- demonstrate a greater capacity to form groups; provide supervision and maintenance; construct, or have water-user grovups construct, irrigation distribution facilities and as additional informattion on thie underground aquifers is obtained.

4.34 _inited States Geological Survey-assisted study: A five-year exploratory program and study of groundwater in the Province is scheduled to begin July 1970 bv the Groundwater Directorate of EPWAPDA with the assistanice of a team of experts from the United States Geological Survey (USGS). The cooperative activities of this group will be scheduled on a priority basis so as to be as helpful as possible in providine data for the above tubewell developments. From a long-range point of view, the USCS stuidv will be important -- in fact essential -- for determination of In overall water balance for the Province out of which will come an understandinv of the long-range role of grnondwater developmnent and it!3 integration with surface-water development. This is discussed further in ChArter 5-

b..5 .. .Q hns hoon in,icated the tubewLs constructed by ADC and PARD utilizing local methods and materials have cost considerahlv less than either the EPWAPDA tubewells or the 3,000 tubew,ells now being undertaken by ADC. Recent experience in West Pakistan indicates

tha!t, Of 83,000 tubeuells inst-alled duringLA~AlE a-pproiatlae.'X4matelyCA the pasLecpat .ea onlv 4,000 have been by public agencies and the remaining 79,000 by private individualsIc Prlvate ftuhoells a-re rcurrently being installedA 4n West Pakistan at the rate of 10,000 per year. The private individuals insta1 ino these wells have been relat-ve-1 large 1nndowners ho have been able to obtain credits from local banks. The wells thev have installed resemble considerably those at Comill- in Easft Pakist-an. Ul.ile thIe stand=_ ard of such tubewells may be far lower than those now being proposed (there . trfouch larer percentage of prouctive wells) nevertheless tn view of their far lower cost (even lower than the foregoing fioiiror. inA i rnt-o if f-nr,1ion oo,r,nnoo la cAo-pricedAcice th (-...4 1t wells involved practically no foreign exchange) and the fact that they have proved to be of success in achieving, agricultural production would indicate a need for further study to determine more clearly the applicability of

indigenous .. ethods of tubewell .onstrucItion in East LPak i4 . Obvlusly, recognition would have to be given to the fact that the number of credit- wo rt hyvindi4v-Aid4aIc fa.rmers int East Pakit 4s 4far Ies, pi-p-----o--ate-y, than in West Pakistan. Consideration should be given to a policy which would glve direct subsidies 'Lo tUhLe privatu sector tLLrough thc Agricultural LIevelop- ment Bank, which would be equal to the subsidies given to EPWAPDA and ADC. - 42 -

Fractional pumps (capacity less than one cusec) and dug wells should also be investigated in this regard.

4.36 Now that several types of tubewell program lhave been successtully launched in East Pakistan, comparative data will become available that can become a basis of more intensive programs. It is recommended that a timelv evaluation be made that will take into account the need for accelerated tubewell investment, rapid verification of groundwater conditions, and coordination of the various tubewell projects. The evaluation should aiso take into account: the organization of tubewell groups and irrigation distribution; tubewell operation and maintenance; and an economic analysis of comparative costs and benefits of different types and sizes of tubewelis (considering also dug wells as mentioned above). Such an analysis must consider not only the need for accelerated tubewell program and technical questions, but also the need for labor-intensive projects in East Pakistan and the desirability of improving indigenous capacity, whether that of private contractors, cooperatives or governmental organizations. It may be possible for the consultants engaged on the 3,000 tubewell project of ADC (para. 4.22) to make the necessary analysis.

Multi-purpose Projects - Feasibilty Studies Available

4.37 As has been described in Chapter 2, multi-purpose (irrigation, drainage, flood prevention and navigation) surface-water projects have been studied intensively in recent years and a strategy has now evolved that is suitable to the small, fragmented landholdings that prevail in most of the Province. The elements of this strategy are:

a. Use of labor-intensive methods (for the extensive earthworks involved) which are well suited to the local conditions of climate and labor;

Experience with embankment construction has been good. In the Coastal Embankment project, over 2,000 miles of 12-ft. high embankments, which provide protection against tidal flooding, have been built by such methods. The Brahmaputra Right Embankment (139 miles long, 12 ft. high), which provides protection against river flooding, also used such methods plus supDlementarv compaction by machine. b. A well integrated project organization for provision of agricultural inputs, including emphasis on rural infrastructure (see Chapter VII, Volume I); c. Adoption of a method of irrigation distribution (i.e. use of natural drainage channels as distribiutaries npus low- lift pumps) suitable to the conditions of small landholdings; and

d. Careful phasing of proJect constructions investments..nd - 43 -

This is practicable since the projects involve relatively small-scale civil works. For example, a primary pumping station (cost $1 to $2 million) can be built in 3 years to serve 1,000 to 2,000 low- lift pumps (50,000 to 100,000 acres). As a second example, embankments can be planned for partial flood protection of a project area or left for a later stage. Careful phasing can reduce the long gestation periods previously associated with multi- purpose proJects down to the point where they will be only slightly longer than the gestation period for tubewell proiects. The multi-purpose projects will moreover provide important flood control, drainage and naviRation benefits.

4=l38 Experience thus far with multi-purpose projects is limited. Only one small project - Dacca-Demra, covering 15,000 acres - has been completed. A second - Chandpur, covering 140,000 acres - is now underway; this proiect has been assisted by two IDA credits totalling $18.25 million (see Chapter 3 for details) Proiect preparation is however almost completed for 8 projects with an aggregate gross area of 1.5 million acres (1.0 million acres net), as follows (see Table 2 for proiect areas and Map 2 for locations):

Darrc Southwest Belkuchi Karnafuli lJpper Kusiyara Muhuri Khowai Pabna Sangu

These projects are technically sound, have high rates of economic return and can be 4ntegrated into a long-ran-e Province-wide nlan for water develonment including flood control. (See paras. 4.174, 5.01, 5.03 and 5.07). The status of prolect preparation as well as the availalhe information on these 8 projects indicates that it is safe to proceed - as a next step in their implementation - to final enoineering followed by tenderina for construction. The financing of final engineering for the first three projects has alr.ea-dy been assisted by !DA credit-s totalling, $3.2 million- Additional financing requirements for capital investments as well as for final engineering are in Table 1. Descriptions of the 8 -ro4er-t are contained in the following paragraphs.

4.39 The cost of the final engineering for the above 8 projects would include in each case revision of the existing feasibility studies. (The cost of preparing these revisions would however be a minor part of the cost of iinal engineeri'ng most of which is for preparation of tender documents atnd design drawings.) As was pointed out in Chapter 2 (para. 2.16), while these stuuies contain much vaLuabUle and i.L factL essential inLf*-Ua.tion, in general they are incomplete in respect to the method of irrigation distribution, economic analysis, project organization (particul arly as regards agricu:Ltural inputs) and project repayment. With respect to irrigation distribution, the - 44 - low-lift pump method (see para. 2.14) needs consideration in each case. As regards economic analysis, the feasibility studies prepared up to now (except for Chandpulr) need to be revised in the following respects:

(a) The true "opportunity cost of capital" should be used rather than interest rates adopted arbitrarily;

(b) World market price trends for farm products should be used rather than internal prices including subsidies;

(c) A distinction should be made between cost of items involving foreign exchange, and those that do not (the former should be shadow-priced);

(d) Time lags to full development should be realistic;

(e) Phasing of expenditures to maximize economic returns should be carefully considered: and

(f) On-farm depuplnnmpmnt And norir,,lt-,,ral inputs n-ed adequate costing.

4.40 As regards project organization, a scheme has now been worked nut fnr rhandni,r (para. 2.17) and smiTlnr QrhPmPe -- although not necessarily exactly the same -- need to be worked out for all future projects. The problem of pro4ect repayment, both as regarsAa noprtioo and maintenance and recovery of capital costs, is an urgent question which must be examined not onl%y on a proetb-r4c basis but cans or. aProvince-wide basi-s; see chapter VIII on the Comprehensive Study in Volume I.

4.41 Detailed descriptions of the 8 projects are contained in the following paragraphs.

4.42 Dzcea Southwest prosect: T,he Dacca Southwest Irrigation project would provide flood control, drainage and irrigation to a gross area of 390,000 acres (300U,000 acres net) located re ar rDacc-a (see rip d)1 Originally, the project included four polders with a total gross area oL 497,000 acres. W of t e hiave bUeen eliminated froLm. the project because of questionable feasibility. Further study has revealed that the two polders 0 0 0 0 in the remaining area of 39 , acres can be best developed by combining them into a single polder.

4.43 The embankments, 167 miles in length, would eliminate flooding from the Brahmaputra River and its tributaries. hne project area is tlooded every year during the monsoon season to depths ranging from 3 to 15 feet. - 45 -

Tvpically, farmers in the area now grow broadcast atis (summer rice harvested in June or Julv) in the areas of medium to shallow flooding and broadcast aman rice (autumn rice harvested in November or December) in the areas of mpd4tm- to depn flooding; both varieties are often nlanted together in areas of medium flooding. These varieties are low-yielding and owing to flooding cannot hbe fertilized- The eliminAtlnn of flooding and the nrovision of irrigation would enable the use of improved transplanted varieties and fer- tili7er andnfnild rpesult in a eriplino of nrpePnt rirp vie1ds.

4.4 F!oI--odlim4nation muiat hi. canrefu1lly nhacsPd sinrc fnrmovra nnnt shift from the broadcast to the transplanted varieties in only one season. This is due not only to the need for farmers to adlust to neo. methods (although cultivation of transplanted aman is well known in East Pakistan) but also because there are logistical problems to be solved (such as provision of seed and fertilizer) and because the transplanted varieties often suffer from water shortages particulariv at the begi4ni4- -a -- A of the monsoon season. For this reason the gradual elimination of flooding iS proposedU whlerebyv thle watSer level inside the polAer WouIl be gradu-6 allv lowered during a period of perhaps five or six years.

4.45 The other major features of the project besides flood prevention WSJUI ltSOLSS~?~,C wvu;'iC eanaLL raiagirigtin.I II ~ . SL)L FSULCordranae, .. I I pi,laIPL) ±15111LY y pu,,pingPUIIIP.Litr stations11LLUI would be constructed at natural drainage outlets along the embankment paral- te 1A C _ _ v_ - IC -1 2______e___e______t_.______Xt= L. n 1I, tllt: i iganipa I . JUIUIOV%I L1J. L[Vt1CiiL UL LLIC llILULd XLL*LtL .ut drainage channels would be necessary to assist in accomplishing drainage.. For irrigation, these same pumnping sLaLions would be operateu in a reverse direction to pump water from the Kaliganga to the interior of the polder.,

4.46 I)uring the months from October through May there is very little rain in the project area and much of the area lies fallow. in addition, there are occasions when the monsoon rainfall is insufficient for paddy so that supplementary water is often required. A primary purpose of the project would thlus be the provision of water for irrigation.

4.47 The density of the population in the project area is very high (about 2,500 per square mile) and tile average size or landholding is verv small - onlv about 2.3 acres. Landholdings are moreover highly fragmented so that the average plot size is only about one-third of an acre.'/ Under

1/ Fragmentation has provided a form of insurance as farmers have tried to disperse their holdings with respect to land elevation. (In years of deep flooding the higher land still yields a crop while the lower land does not. The reverse is true in years of shallow flooding.) Land consolidation to correct fragmentation will be encouraged by the project as it will largely eliminate the flood hazard. Land consol:L- dation will also be encouraged by the availability of irrigation; see para. 2.14. - 46 -

these conditions the provision of a conventional gravity irrigation distribution system is impracticable. This is further complicated by the extreme shortage and high cost of land in East Pakistan. Fortunately, however; it is possible to accomplish distribution of irrigation water by making use of the numerous natural drainage channels. Water can be pumped from these drainage channels by means of small portable low-lift pumps of about two cusec capacity.

4.48 Another important benefit of the project will be the provision of intrrior navigation- Rural roads are lackinQ in the nroiect area and moreover the economic level of the typical rural family is too low to be nble to afford wheeled vehirles; native country boats on the other hand are available everywhere. The improvement of the numerous drainage channels will provide4 mny new navigable waterways and tfhe maintenanrce of a minimum water level inside the polders during the dry season as well as a .o-4a lve-l d-uJ4nn tho wot- cr..nn4 411 glreatly etndr-aanIlabl the th of navigation. At some of the principal natural channels, navigation locks wouldA e proviAA4 to perm it boat t- antear and leave the empoadered areas.

4.49 Accordin- to the existino feasibility stud-, construction of the major works would take five years; the main construction item would be 167 u--4les of em.,ankment. The 135-mile e.mbankment for the Brahmaputra pro4ect took five years to build although very little work was done during the first two years. In view of the expea4rl ---- a -4A it4nd r,A beku hosibe to construct the project embankments in four years. On the other hand, even if a somewha,t longer time is req,,4re, the --o-noc- 4ustictlon of the project will not be in doubt. This will be due in large part to the phlIas ing oil the .onstruction tlOhat i0s planned o0 as to obta4n interim benefits, as follows:

The first item of construction will be a primary p Ling station 'in tlhle northern (upstiream) part of the project area. This station will be similar to what was uescr'Lbeu iLn para. 4.371(d) asbove. .IrrigatlonLin thLILs part of the project area during the drv season will thus be available possibly even before flood pzCveutionL has been accomplished. If confirmed by detailed studies, the other 7 primary pumping stations may also De completed and operated in advance of the availability of flood protection.

Construction of the flood embankments would be started at the upstream end and advance in a downstream direction, along both the Brahmaputra and Kaliganga rivers. During the second and subsequent construction seasons temporary closures running across the poiders would be provided for interim flood protection to the extent feasible. - 47 -

These temporary closures need be only about six or seven feet high (onlv about half that of the main embankments) and they can serve later as permanent rural roads similar tn thnnp hbins constructed throuehout the Province under the Rural Works Program. The temporary closures would nrnvifiz nprntPrtrinn aA'nlt ahntit a two-veAr finod An compared with the main embankments which provide protection anninsr a lfOfl-ypr flond and are mnreoner rnneprvntiv-lv designed to withstand wave action, settlement and erosion from various causes. Whilo runoyear protection would be insufficient to enable farmers to shift from broadcast to transplanted eman it .uld be enough to ernable the.. to grow a greatly improved transplanted aus crop. This crop is hkarvested in June whereas the peak of flooding from. the Brahmaputra does not come until late July or August.

As construction of the embankments advances further, practically comi.plete protection will progressively uecorue available. Thus almost complete development in the upstream- part ofj thle po'lder wJl be-vr posil I-___ be-or itL ~LL~dW~dLLL LLL~IJU±UL WL.± UC FU%.pUo.LU CVULI Ut:LULC LL is entirely closed. i/

4.50 Costs of the Dacca Southwest project are sufficiently well-known to inrdicate that the unit costs per acre are likely to be smaller Ltan for the Chandpur project. Since cropping patterns and economic and social candi- tions are similar to those existing on the Chandpur project, the benefits and therefore the rate of return should at least be equal to and probably exceed that of Chandpur which has been established at 18.5 percent. L/

i; Aside trom construction phasing as described, it is not practicable to subdivide the polder so as to have an initial project of smaller size. There would be serious local opposition to such a plan, first by the large number of landowners located in the right-of-way of the additional embankment that would be required; secondly, by landowners in the area excluded from the polder. Thus it would not be feasible to get agreement on the limits of the reduced polder.

2/ The appraisal report for Chandpur shows 22 percent. However, an adjustment for "sunk" costs previously incurred reduces this to 18.5 percent. - 48 -

4.51 Consulting firms have been engaged by EPWAPDA to revise the feasibilitv report of 1968 particularly regarding the matters mentioned in para 4.39 but including also a review of flood control features (embankment design) and design of numning station intakes to minimize sedimentatinn- The consultants will concurrently prepare final designs and tendering documents for initiatinn of ronntrurtinn- Tt i8 antiri1 tApd that thp fenaihiliru report will be completed by September 1970, and that construction can be init-iate- hu TJaniuarv 1971= Construtiotin should hp cnmnlptpd in four or five years at a cost estimated at about $60 million (total equivalent). This estimate was based upon preliminary information obtained from thle 1968 report, and subsequent information obtained from studies now in prog.res-s.

4.5 Th-e Karnafuli4rr,4t includes four sub-proJects of which t, are proposed for early development. These are the halda Unit with a gross area of 131,500 acres and the Ichamati Unit w'th a gross area of 16,"0° acres. Feasibility reports on these two Units, botlh dated May 1968 were prepared b-y the consul-Ing er.gineering fi.., Just-in=Court-ney=Hohlwe^g-Watts. A- nIA uyLLL Lt*t LA.. LJ . ,±.cL iS fI 5 L.lt'JL-I~ Lt ZJIWF rY Iu I C rttu -.t credit of $2,400,000 has been made to the Government of Pakistan for revision of the Karnafu4l and Muhuri feasibilty studies and fo, the preparation concurrently of tender documents and final designs. The scope of work under thLte credit allso includes a reconnaissanca of the Co.iillaNoakhali and Little Feni project areas which have a total area of over 1 million acres (see dCetail.ed3 dXiscUSs4on later in this chapter).*ih.e work to be carried out under this credit thus comprises a major step forward in the development of the4 Southeast region, .1i eL o&A LJL&LAa -

j~~~ TL J I in w1_tIc __tno I 4 ty4*.._-..v..Cok .4 J* Vi L ± theLLe C dy Le a s 'L U -L -.LLL LUUy Uf LIth ALa.L la UlnLt oni1tasakt much useful information, additional surveys and studies are needed before final desLgn and construction can be undertaken. The report pro-poed that flood protection be provided by a dike system with a total length of 70 miles. There would be eight miles ol dike with an average height of about eight feet along the right bank of the Karnafuli river. There would be 26 miles of dike along each bank of tne Ralda river witI an average height of about 10 feet and with a maximum height of 12 feet. Finally, there would be 10 miles of dike averaging about eight feet in height along the two principal tributaries, the Sonai and Sarta Khals. These embankments would eliminate flooding from the streams mentioned.

4.54 Flooding occurs every year during the monsoon season when large portions of the area are inundated to depths of three to six but as much as nine feet. Typically, farmers now grow aus followed by aman rice in areas of medium to deep flooding and aman rice in areas of shallow flooding. 1/

l/ Aus rice is planted in Marchi and harvested iLn J.ue or Jly. L Mn rice is harvested in November or December. The aman rice in the Karnafuli area is of the transplanted rather than the broadcast variety. 4.55 Since flooding does not ordinarily last more than a week or two, tAC farmers gr.tow transplantdahA . than l.o-e-oriebAdno rice. On the other hand, the transplanted rice is often severely damaged by flash I lood s. i.ce yields, which are presently low, could be about tripled through elimination of flooding and provision of irrigation which would provide the biasiLs .or amodern.ized agriculture.

J5L~ U A- UU ILLU a ofL.LLLL

*t .J 4.56 AaproposeIiU% propose' in thIe A968.'JJ reportreotC L L th e eu-.b"ankedCM.ULm~e&IC - chan.nel.a,LC.S.A of theIL.f Y.aldaASU. A.fl ~ 1 river would have a discharge capacity of 22,000 cusecs with a three foot

'reeboard. TIe mL-im-A-±1I L.LooAU A.QlW observeA Lto UasLe hlas bee. ounly 12,750 cusecs; however, the consultants estimated that the higher capacity would De requireu because confinemenHt of the rLver by embankaents _l natural valley storage and therefore increase the discharge as compared with natural conditions. Tne current stu±Les will reLvjew the consultants' analysis particularly as regards scouring and resulting maintenance problems.

4.57 The other major features of the Halda sub-project, besides flood control, would be drainage, irrigation and navigation. For drainage, tne consultants recommend excavation of a nine-mile long channel, called the West Main Drain. Some additional drainage works might be necessary for the large low-lying area in the southeast portion of the project and this is being studied under the current engineering project.

4.58 For irrigation of the Haida Unit, the consuitants proposed a main pumping station (capacity 1,750 cusecs) which would lift water from the Karnafuli river into a main gravity canal at elevation 23 feet. After flowing in this canal for a distance of five miles, water would be relifted by a second pumping station (capacity 1,060 cusecs) to elevation 45 feet. The irrigable area would be 84,600 acres net and distribution would be by means of conventional gravity-type canals. The system proposed, besides being costly, would not be suitable under the conditions of small landholcd- ings that prevail. The average farm holding is about 2.6 acres of which 2.1 acres are presently under cultivation. Conditions in this respect are similar to those at Chandpur where the average holding is two acres which is fragmented into as many as ten pieces. Also, as at Chandpur, the project area is crisscrossed by many natural watercourses (khals) and excavated drainage channels which are tributary to the khals; there is thus hardly any point in the project area more than a mile from an existing drain. It: should therefore be practicable to irrigate a large part of the area by low-lift pumps taking the water from the khals and thus eliminating the need to take land for canals. (This is the system proposed for the Chand-- pur project.) The water would be pumped from the Karnafuli River and uti-- lizing low-lift pumps it should be possible to irrigate a net area of about 55,000 acres located between elevations 7 and 25. The maximum static lifl: would be 18 feet and the average would be only about 9 feet -- much less than required by the pumping stations proposed in the 1968 report. Follow- ing further investigations it might be possible to extend the estimated area beyond 55,000 acres, either by additional pumping (relifting) or by development of groundwater, particularl- in the higher foothill areas. For the latter type of development, groundwater surveys and explorations 50 - will be necessary but these will be carried out gradually without holding uD the main. first stage development; the surveys and explorations will however be included as part of the current feasibility study revisions.

4.59 Still another important benefit of the project will be the provision of interior navization. Rural roads are lackine in the nrniert area and moreover the economic level of the typical rural family is too low to be able to afford wheeled vehicles; native countrv bonat on the other hand are available everywhere. The improvement of the numerous drainage channels will Drovide many new navigable waterwavy and the maintenance of a minimum water level inside the project area during the dry season AR well an a maximum level during the wet Reamon will grentlv exvenA tha availability of navigation. At some of the principal natural channels, navigation locks would he nrovided to nprmit honat tn enter anA leava the empoldered areas.

4.60 The 1968 feasibility report on the Halda Unit of the Karnafuli nro4ert eontAina muicth tsaefiil informatinln nnprtriculnrlu with reoard to the -- ~ -,~ ~ ------J _ 1 flood control embankments which constitute the most costly project feature. The conts of other mainr nro4ect wnrks (maior drainnage rhnnnels, pping stations, regulators and navigation locks) will require revision taking into accotmt the alternative approaches and designs described above. Of these the most important revision concerns the method of irrigation distri- buiotAn with4n the pro4ect area. The rertnf- oin t Hlolda Unit is also in= adequate with respect to project organization (see para 4.40) and project re-a-yment. Reg,ardinrg the l1-atter, an analyi4a ia needed teai r payment capacity, water charges and the time period over which they can 1b AprogOreively 4 MpS-A 11/

4.61 ,.C.,S4on of the Y,alsA feasibility report, as described in thUe preceding paragraph, will be needed before appraisal by the Association of a cor.struction creudit can be unAertaklen, On the otAher hLand, thLe re0La- tively advanced status of project preparation as well as the favorable judgment regarding economic _ustiication indi4cate that the Ralda proj ect j %A L. , L^vLu,^ . .L J.,LL t_ L.L L U JU LVL. IssIL ol *Eu L cX can proceed without delay to the stage of tendering and final design.

1/ In the case of the Chandpur project, these subjects were studied in a preliminary manner by the consultants. The province has agreed to prepare within 18 months a plan providing for recovery of operation -.d m-aint-enance (O L MI -ot of t#-be w.a4 n .- ^4 n4 ,t-rk, estim ,ated n. oa...... l '- - -, .- , - -L... rn J- & -J-ttj -A V0 OUth-,. at Rs 15 per acre. 0 & M of the low-lift pumps (including fuel), esti- -ated at LD0 .7 per acre, will' bue collected lrom fza,uers on a sca0'Le riLs- ing progressively to full value in 10 years. Recovery of capital costs are recognized to be more difficult; the Province has hwv a8e- U-L-LLt--L-,I.&= L V-L1%C L IM&UWCVCL 0.5LCrtU to prepare a plan for recovery of as much as practicable of the capital cost. ('See proJect grCeement signed with IDt,A, Chandpur II irrigation Project, May 15, 1970, Section 2.07.) 4.62 By proceeding now with engineering for the Kernafuli project (Halda Unit), it would be possible to initiate construction by about December 1971.

4.63 As proposed in the 1968 feasibility report the Ichamati Unit wi:Ll have a gross area of 16,000 and a net area of 11,800 acres. In 1969 it was noted that 6,300 acres or a little more than half of the irrigable area were already irrigated by low-lift pumps. Water supplies for these pumps have been made possible largely by two regulating structures already constructed by EPWAPDA. These structures also prevent flooding from the Karnafuli river. The project area is still subject, however, to flooding from the Ichamati river.

4.64 The! local farmers are anxious to expand their successful irriga-- tion and believe they have sufficient water to install additional low-lifl: pumps. Interest was expressed in a 100 cusec high level canal from the Karnafuli river similar to that proposed in the 1968 feasibility report. However, as an alternative, additional development of the Khals, as has already occurred, including utilization of the Ichamati river as an irrigation distributary would be considered. Groundwater would also be considered. As a matter of fact, local residents have already made applicatiLon to ADC for 40 tubewells.

4.65 With respect to the Ichamati Unit, extensive revision of the feasibility report will be required as has been indicated above. Tender- inp the final desien for this small proiect should be deferred until after completion of the revised feasibility study.

4.66 An initial determination of the economic justification has been m4ade fnr the Knarnafuli nrn9ert (Halda Unit) based upon a comparison witth the Chandpur project. This comparison revealed that costs per irrigated acre on the Karnafuli pro4ect will he less than rhandpur, and that benefits will be greater. Thus it was concluded that a rate of return for the project Is likely tn exceed that comnilted at 18.5 nperrent fnr Chandpur=

4 4 4=67 Twn conniulting firms (one Pak4itanland one exnatratenas ir!t venture) have been engaged by EPWAPDA to perform final engineering and revise tho fea4h4slb ty vrpnort. RBasl Imn piivrant orhadiwlee 'hreovIse feasibility report should be completed by mid-1971, and making use of the final e.nineer.ng work, it should be nnpo4hle to init4ate constrci or 4 in January of 1972. Preliminary information indicates the project will cost I,' .1 4lli Cor.st.ouruci should be r.pleteA in about four year

4.68 Muhuri 2 rojet: The gross area of the project is 101 Ann acres. A feasibility report dated January 1967 was prepared by the consulting engi- eering fir...Techno=-Consult of KaraCh4, 1Lhore an.d Dacca. W.h.ile Ithis feasl= bility study contains much useful information, additional surveys and studies will1 b-e r.eeded before fiC-Ia'lesigS,. tL.L.A. LhUe m U Los A.a.Lconern an'uthmetLho cor.struction oiL.Lr igcan b-eA un.dertaken..is trib L i 0,.II. these the most important concern the method of irrigation distribution - 52 -

within the project area and drainage. The report is also inadequate with respect to project organization (see para 4.40) and project repayment. Regarding the latter an analysis is needed to establish repayment capacity, water charges and the time period of which they can be progressively im- posed (see paras 4.39, 4.40 and 4.60). As previously mentioned, an IDA credit of $2,400,000 has been made to the Government of Pakistan for the revision of the Karnafuli and Muhuri feasibility studies and for the preparation of tender documents and final designs.

4.69 As proposed in the 1967 report, the project would involve an expenditure of Rs 36.5 million including interest during construction or Rs 34.7 million if the latter is omitted. These figures were for a proiect including a limited amount of irrigation, in an area of only 8,700 acres, with water obtained from the Muhuri river by pumping and thence through a conventional gravity canal system. The small area irrigated was due to the limited water supply available from the Muhuri river during the dry season- The report also describes a project limited to flood-control and drainage only. This proiect would have a cost including interest during constructi-inn of Rs 33.0 million or Rs 31.5 million if the latter is omitted.

4.70 For flood control there would be about 92 miles of major embankments (crest width 14 feet, average height nine feet. freeboard above design flood 3.5 feet) including 21 miles along the right bank of the Muhuri, 19.5 miles along its left bank, 20 miles alone each bank of the Selonia river up to the point where it discharges into the Kalidas Khal (an existing man-made channel) and 12 miles alona the right bank of the Feni river. In addition, there would be 56 miles of smaller embankments along some of the tributarv streams.

4=71 The exnenditure of Rp 4.8 mi11ion pronosed for channel improvement is mainly to eliminate bends in the Muhuri and Selonia rivers so as to increase the4r carrying capacities. With these i^rrovements plus the embankments, the Muhuri river would have a discharge capacity of 25,000 cusQecs, The consultants hnve estimated by synthetic methods that such a flood would have a period of recurrence of 10 years. Such a degree of protectio.n seems reaso.nable amider the conditions. prevaili4.g. TI..e 19617 report contains no discussion on the velocities in the Muhuri and Selonia channel's AS related toaheir stability and the resultling maainten,ance problems. The width of channels provided seem ample to keep velocities low. Nevert-heAless, thne flood .hydrology anr.d hydraulics of th'e 11-1hur' an' Selori'a LAL~~~~ LJ.U1.JU A1y1.1L1JL1.J~~~~~~~~~~~~~~~~~y%. L.)tU.1±U.I[1iuOLLU LI~~~~~~~~UL0UJ~~~~~LL~~~ ~~~ ~ ~ UL LII~~~~~~~~~IU1LUL J. ~~~~~~~~~LIUC embanked channels will be reviewed.

4.72 The Muhuri river embankments would terminate at the Dacca- bnittagong nlghway bridge near Haripur which is about 10 miles upstream of the mouth of the Muhuri river. The consultants felt that the Muhuri river downstream of the bridge need not be embanked thus allowing spill to occur into the portion of the project area between Haripur and the outfall. It was thougnt that this zone would act as a detention reservoir. - 53 -

However, there may be two reasons against such a plan. First, with the improvements upstream, there would be more flooding in this area than at present and this would probably not be acceptable to local residents. A second reason is that with the banks of the Muhuri river in the ten-mile reach downstream of Haripur quite high, it should be relatively easy with embankments no more than a few feet high to carry most of the flood flow to the mouth of the Muhuri river.

4.73 The report indicates an additional embankment is needed to form a closure between the existing right-bank Kalidas embankment and a coastal embankment at Sonapur. The cost of such a closure and whether this would be part of the Muhuri proiect will be explored.

4.74 Although the 1967 report contains some discussion of the drainage problem, this will need further study first with respect to the low-lying area downstream of Haripur that has been mentioned above. Secondlyv furthLer analysis is needed with respect to other low-lying areas that cannot be drained during high levels in the main exit channels- namely the Muhuri Ard Selonia-Kalidas. It seems probable that pumping for drainage would not be economically -iustified. On the other hand- further knnwledge is needed regarding the frequency and depth of flooding caused by inadequate drainaae due to lack of exits AlsRO suitable cronping patterns would have to be adopted for such areas. Thirdly, the drainage system should be designed takInp Into account its uttlization for the diqtribition of irriaat-ion as is discussed further below. Finally, use of the larger drainage channels for naviont4ion will hp eonsidprpd as was fournd to hb feasible nt the rhandA pur project where a highly important additional benefit was obtained. The nrovision of nauigatinn icB considered to hp of oreat aig nf4ianc-a *nAer th conditions that prevail whereby rural roads are lacking and are likely to be comparatively costly. Moreover, wheeled vehicles are scarce while country boats are plentiful.

4.75 The irrigation plan proposed in the report would provide irrigation by gravity to only 8,700 acres in the higher, northern portion of the project area. The small area irrigated would be on the basis of the avail- ahi F Arv-Qeason flo omf^f the Muhurl river. A conventioal gravity system is proposed which under the conditions that prevail (small landholdings)

4.76 The report proposes sluIces at the mouths of the Muhuri river and Kalidas Khal to prevent flooding and saline intrusion from high tides. The design *o.,::ld be sim.ila. to that of th ex4stir.g EPWADA regulator across the Little Feni river whose drainage basin adjoins and is west of the

M Whuril river drain.age basin. T.AL.is regulator has bI-een successfLu'l in eliminating flooding from tides and saline water intrusion and, in addition, has in the past few years ena,bled the lrrigation by means of low-lift pumpiu - 54 - of about 5,000 acres. This area could be further extended if the water level upstream of the regulator could be raised but this requires some flood control measures upstream. 1/ (See description below of the Little Peni proJe*.4

477 A. of Maxy 1Q9Q 113 lnw-lift- pps had been installed bu ADC along the Muhuri and its tributaries and the available water supply is ample fo: some additional . The area irrigated per pump is about 45 acres or 5,000 acres total. It would seem that, as a next step, im- poundmer.t of water by the proposed Muhuri and Kalidas sluices (as at the Little Feni regulator) would permit further installation of low-lift pumps.

4.78 According to the area-versus-elevation curve in the 1967 report, there are 68,000 acres below elevation 25 and 11,000 acres below elevation. 15, the difference being 57,000 acres. The bed of the Muhuri river at narLpur i8 at about elevation 6 whAereas the existing banks are abouteleva- tion 20. The same elevations pertain in the Selonia river from its mouth at the M"uhur' ri-ver to a pointil several iles upstrea. T.he banks of each of these rivers are higher than the surrounding terrain. The high-water level proposed at each of the two major sluices (tMu-Ar and Kalidasl is elevation 15.5 (during the monsoon season). If a level of about 12.5 could De ma'intaiLneued 'ur'Lng theld.Ly season it s.Lould be possible to sere almost all of the 57,000 acres. (Data are not readily available to determine the corresponuding net area. Assumuing proportionallity with. thwe K.r.nafulJi (Haldi Unit), the net area would be 42,000 acres.) Water would be distributed by meaxw of low-lift p-ups obtaIlning their supply from the drainage system. Since, however, the flow of the Muhuri river is insufficient during the dry season, another source oUL water wvould be required.

4.79 Import of water to thi'e '-ulurL proJect area, on the basis of available information, appears practicable. The source would be the Dakatia river wnich receives water at all tilues f rum the MYeghna. From aerial reconnaissance and examination of available maps, it can be observed that the Dakatia has a wide navigable charnnlel as far inland as Rabiganj (14 miles east of Chandpur). The Dakatia is also navigable part of the year as far as Laksham (17 miles east of Habiganj - see IWTA navigation maps). Laksham is at the drainage divided between the Dakatia and Little Feni rivers. In view of the generally very flat topography it should be feasible to transfer Meghna water via the Dakatia to the Little Feni. Works needed to accomplish this might include a pumping station cum navigation lock on the Dakatia river. Obviously such a pumping station might ultimately have to serve a large part of the Comilia- Noakhali area (over 800,000 acres gross). However, such a pumping station

1/ The Little Feni regulator has 20 vents each 12 feet in diameter. The reglator is capahle onf mnnint-aining a water level of elevation 14 feet but upstream residents have objected so that the normal level reached (Irn October-November) "^w is elevation 12.5 feet, Low-lift pumping draws down the impounded water to about elevation 6 feet (in February/ "'s*c'', - 55 -

could be developed gradually in stages. By making use of the existing rvul1ator at the mouth of the Little Feni river as well aR of the e,eiRt- ing channels 1/ that connection the Little Feni with the Selonia it should be nnposihl t tr-ansfer water from the Little Feni to the Selonnia/Muhuri river system. Possibly some raising of the crest of the Little Feni regu-. lator alono with onme anillarvy wnrks (low dikes) might be needed. To determine the practicability of such a scheme additional field informatiorn, part4cularly with regaard to the pvrfi4 lea and cvrosas-rcti4o.a of the n.atural stream channels involved, are being obtained. (It will not be necessary to get comlete tpography of the areaadjace4nt to these st_eam channels. Actually most of this terrain is reportedly already mapped at a scale of four miles to the inch w4th one-foot contours.) 4. 580 AA. report pr-epared by TI 4rn.J.ue 1'O'6. en,t4i-led "Grou ster 4n fl*jJJ L. & t.~ J -Lp *S .1 6-A1~V "~WC&% C L* East Pakistan" by H. V. Peterson, groundwater hydrologist, indicates (pages r57 a.4 58)% that the prospecits Adevelop-ing grounc-wter seem favorable or,i .JI WL~ .JJ -LA L. L.&& J1Jp..L~ ~ ofJ ~ LJpL LL L u waAWG %_a L uLUL~U the east side of the Muhuri valley but much less so on the west side. Since LtbUUbLels W-U.LU provide water fLor iLrrLgatiLon ofLhi.Lgher Lands (abo-ve eleva-- tion 25), groundwater investigations will be included as part of the future feasibility study re-visions. 41 Af h Deen indicated auove, rav'gat'on i8s an aspect needing further study not only with respect to the Muhuri project area but also in connectlon with tne possibility of proviiuug navigation "lnks witn tne Lii:tle Feni and Dakatia rivers.

4.82 An aspect of the project that is scarcely treated in the 1967 report but is vital to success of the project is that deaiing with organization. Although thus far only EPWAPDA has been carrying out the planning of this project, other agencies should be directly involved, even in the planning stage. Basically a strong project organization is needed headed by a Project Manager. The agency to whom the Project Manager would reporl: is a matter that would have to be settled by GOEP. Personnel of the various agencies concerned should be seconded to the Project Manager and would form part of the project organization.

4.83 In the note on the Karnafuli project a preliminary determination of economic justification was made. Such a preliminary determination is not now possible for the Muhuri project because insufficient information is available with regard to costs of the three primary elements: flood control, drainage and irrigation. This will, however, be done by the time the proposed interim report described below is submitted. With the infor.- mation developed in the 1967 report, as supplemented by the additional flood-control studies described above the cost of the flood-control items

1/ EPWAPDA maps for Feni Sub-division, Noakhali District indicate two connecting channels: Gazaria Khal connecting Little Feni river and Kalidas Khal; and Madhupur Khal connecting Little Feni and Selonia rivers. - 56 -

should be possible of sufficiently accurate determination after perhaps a few months of work. The draina8e requirements seem more complicated and probably would require a longer time. The irrigation costs will depend on the practicability of importing water from the Meghna river via the Dakatia and Little Feni rivers and for this still more time would be needed; see further discussion below.

4.84 As regards benefits, the available information indicates that these would be relatively high in the Muhuri project area. Of the gross area of 101.000 acres, 72,500 acres are now cultivated, the nrincinal crops now grown are as follows:

Aus/T Aman 36,000 acres T Aman 20-000 B Aman 6,000 Aus/T Aman/Rabi 5-400 Miscellaneousl/ 3,100

Total 72 500 "

4.85 The present average yield of the kharif rice crop is only 12 maunds per cropped acre but it is 19 maunds r acre latter is about the same as at Chandpur. The total area now cultivated

durir.g'.S~..~fl the rabJLC&.L seaon..ontO~aatow to onlyI 10,30 acre (at Chndu it 5 '~t AuJ.l.L tJJ& 4.w ,.415Y CZL. C %I ~.LSCLLLLUP 4- UL .1.L .LZi 44,000 acres). With complete water control it thus appears that the benefits at M.IU..-4 -WtJU.LU U bstAaLntiLLJ.A..Ly 1i.L?;ILCL LIh&aLnL atChanLUdpur. OIn thLe cost side, while information is deficient at present with regard to the

pLMLaLy CL:UICUL.M -- .LUVUU UULLLLUJ, ULd.LL1dgt dIIU LirgaLLUn, lIn VJ.LW of similarity of terrain and size of landholdings, it seems reasonable to ex-pect thlat costs per acre will 'De s±ini.-ar to tilat o0f thle arnafuli project. These costs do not however include the cost of importing water via the Dakatia-Muhuri water transfer scheme. Since however this scheme would ultimately benefit substantial areas in the Comilla-Noakhali and Little Fenl areas, the allocation chargeaDle to munuri would nor nave a serious effect on the rate of return of the Muhuri project particularly if the development of these areas proceeds as expected (see below in this chapter). In general, therefore, the outlook appears favorable as reagrds the economic justification of the Muhuri project. Substantiation of this preliminary conclusion should be possible at the time that the interim report (see next para) is issued in August 1971.

4.86 As now scheduled, the feasibility study revision would be completed April 1, 1972. An interim report setting forth the principal dimensions of the major project works would be completed August 1, 1971;

I/ T.I.udes 2,700 acres of rabi orly. - 57 -

this report would also reach preliminary conclusions regarding the Dakatia- Muhuri water transfer scheme. Tender documents would be prepared beginning August 1, 1971 and these would be completed at the same time that the feasibility study revision is completed in December 1972. Tendering and preparation of final designs would go ahead immediately and construction could be initiated January 1, 1973 with completion 3 years later. Based upon preliminary data, cost of the project has been estimated at $12 million.

4.87 The Pabna project: The Pabna project is located along the left bank of the Ganges river between Abdulpur and its confluence with the Brahmaputra-Jamuna river; see Map 2. The project would provide flood protection and drainage for a gross area of 453,000 acres of which 316,000 would be irrigated. The project is being studied by a joint venture of two consultinR firms. 1/ The consultants are considerinR various alternatives that have been suggested to them and it is expected that the report will identify the most practical plan of development.

4.88 Flood prevention would be provided by an embankment 188 miles long encircling the project. The embankment along the Ganges would be 49 miles long. Drainage of internal rainfall runoff during the monsoon season would be accomplished by a system of drainage channels utilizing the existing water courses as much as possible. The accumulated runoff, to the extent economically justified, would be removed by four pumping plants.

4.89 Many schemes have been considered for irrigation of the area. The one that the consultants nnw seem to be favorine would involve a main pumping plant on the relatively stable Karatoya river at Bera where backwater flow is received frnm the Rrahmanutra-Jamuna_ The Raral river would be used as a conveyance channel as would two rivers in the interior of the nrnipr'r the Trhamnti and the Atrai. The ronniiltantsa ar alon ,-nnaider- ing use of existing drainage channels as irrigation distributaries (supplemented by low-lift pus) as is pro,posed for t-he Chand-Apur s-nd Dacca Souith- west projects. It has also been suggested that groundwater might be an imnortant water Rnhtre-e nartir,ularlv in the western nortion of the nrniec-t area.

4.90 Phasing of the project is a crucial matter in view of its large size. Finanr.cial anrd orgalnizatiornal constrairnts nreeds are i-ortant considerations in selecting a first-stage project.

4.91 EPWAPDA and its general consultants are keeping in close touch with the pr-4ai consul t-nts- tons a'e thttai i-hbefe 4 1lit4r44-r,_i p 4 adeo- quate and that sufficient information is provided (for example) as regardEt 4 the abv4ssuesi-~,naionanAi-none*-ln* ~h.. ar.d to see t-hat adequate0 -. treatmer.t-.n--,a. 44is4- a gie.-t.. -W..an.4 rJc organization and repayment capacity (see paras 4.39, 4.40 and 4.60). The feasibilit-y report is eected to be avail.'able irn M4d1970. he C08t of

.1/ Sar.yu Consultants International, LInc. of Japan andU AsociLatedu Architects and Engineers Ltd. of Pakistan. - 58 - final engineering (preparation of tender documents and final design drawings) has been estimated at $4.0 million total equivalent; it could begin January 1, 1971, if financing can be arranged, and completed june 30, 1972. Con- struction of the first stage of the Pabna project (gross area 220,000 acres, net area i20,000 acres) could begin in about September 1972 and would require 3 or 4 construction seasons to complete.

4.92 Preliminary information indicates that a rate of return equal to or exceeding that for Chandpur (rate of return 18.5 percent) will result for this project. Based on information currently available, costs of Phase I of the project (gross area 220,000 acres, net area 160,000 acres) have been estimated at $50 million (total equivalent).

4.93 The Belkuchi project ares is bordered on the right bank by the Brahmaputra river and on the left by the Karatoya river in the Pabna Dis- trict. Gross area of the project is 101,000 acres, 70,000 acres of which are proposed for irrigation. The Brahmaputra Embankment completed in 1968 protects the area from overbank spill of that river. However, the area is still vulnerable to flooding from the Karatoya-Hurasagar river and Brahma- putra backwater to an average depth of five feet. While there is always an abundant supply of water in the adjoining rivers, most of the land remains uncultivated in the winter months and agricultural yield is poor due to floods and droughts. A feasibility report in draft form dated September 1968 prepared by the consulting firm Techno-Consult of Pakistan is available.

4.94 Thirty-five miles of embankment are proposed for flood protection. Three pumping plants would supply irrigation water, and two of these plants would also serve to drain the area during the rainy season. Existing channels coupled with one sluice are planned for drainage, and the gravity irrigation plan developed by the consultants would require 58 miles of main canal and 101 miles of laterals to convey water to the irrigable lands. The water source proposed for the project is the Karatoya river which is said to be reasonably stable.

4.95 Since the area is traversed by numerous channels, it is believed that low-lift pumps might be utilized to convey water from natural waterways to the irrigable land and thus avoid the right-of-way problems and expense that would be encountered with a gravity system. Small farm holdings averaging two to three acres in size would make right-of-way problems acute on this project. Use and improvement of the natural channels would aid navigation by country boats, the only mode of transportation available to the greater bulk of the farmers.

4.96 Representatives of Techno-Consult said that tentatively they thought that aiX to eight months would be needed to studv the low-lift pump alternative. While the economic analysis made in the 1968 draft report is hbttar than thnt in many other repnnrt in that disr-unting procedures were employed, further improvements in the analysis would be reeded; see para 4.39. - 59 -

4.97 The project appears to have good potential for development and costs should be reasonably low. considering that much of the needed flood prevention works already exist.

4.98 Organizational considerations are as for the Pabna project; see discussion above.

4.99 Baned unon comparison with other proiects (such as Chandpur) it. appears that the project should have a favorable economic justification. Assuming that the same consultants that made the feasibility report of September 1968 are engaged, that their work is closely monitored by EPWAPDA and its ognerA1 rcnsutltants and that the feasibility study revisions are completed by the latter part of 1970, it should be possible if financing can eh narvAane tn arqrt rniatriurtinn of the Relieirhi pro4ec-t in about September 1972. The cost of the project is estimated to be $24 million based upon currently available data. The…cost of final enginering hAs been included with that of the Pabna project which is adjacent.

4.100 The Upper Kushiyara project is located in the northeastern corner -. ..1..p C ..11efa 1i4S.-i-+,4 Ai.lafla-rF~1het y Ton,n A-or4pij1 ti-tn1 ofL I - -src,90------about 20 miles east of Sylhe Tn production is considerably below what it could be if damaging floods were prevented, and if water could be m-de available to crops durin. the dry season. Improved drainage is also needed.

4.101 The Sylhet District is in the heaviest rainfall area of the P'roviLnce. Ih e average annual ra-infall is 'Kr inches in th11e project area. Annual extremes range from 104 to 422 inches annually. The dry season,

LNiJV M_LUL 4II iJU_i 1ILi, JO bA.. uA..U LWL MVJiLL OLPJL L^ i1& A~J U k_ Ua 'i 'NvmVberIC: L' rughs March, is n*D m."l #CJ^awo monthVI shor rlV ICh,[an mUos *art-s of the Province - a significant factor as regards irrigation requirements (see below). 4.102 Hea-vy rifll and cocret. flood flo-ws in the rivers cause LiJ4 eiavy rana.LI".J an'u concurreniLL ..LJUiJA LLW Li L.L1 L.VLaOIO extensive flooding from direct rainfall, reverse flows along the drainage

19IAGhaOl,anAu ovecrban. spills' from thle Oi. and, SUmaL rivers. M54' r-- ect area is, however, above the influence of backwater flooding from the

4.10 ThViIle prolJect area s.Lopes fLrom abuout 7I0 feet abuove sea 'level at the upper end, to 35 feet at the lowest. Farmers in the area say that the Kushiyara river overflows last from 7 to 15 days, that owing to inadequai:e drainage fields remain flooded 3 to 5 months a year and that severe crop damage during the past 14 years has resulted from an embankment along the Indian side of the Kushiyara. l/ There have been some local Thana counc:Ll efforts to reduce flooding and three khals have been closed to stop early flooding from the Kushiyara.

I The Kushiyara embankment would be about 25 miles long. The Kushivara forms the international frontier over about half this distance. - 60 -

4.104 A feasibility report dated May 1966 was prepared for this project by a study team made up of EPWAPDA employees under the general supervision of International Engineering Company. The 1966 report recommended a project covering a gross area of 87,000 acres. Flood protection of 78,000 acres would be provided by means of embankments; about 50,000 acres would be drained by gravity; and 40,000 acres would be irrigated by pumping from the Kushiyara and Surma rivers into a conventional gravity canal distribution system.

4.105 As the 1966 report brings out. prevention of flooding coupled with drainage are the primary needs of the area. The report recommends protection against a 100-year flood; however a lower degree of protection - against say, a 10-year flood may be more economic. The resulting lower embankments would not require as much right-of-way and would be less of an impact on the villages that occupy much of the higher ground adjacent to the rivers. Drainage outlets. acting as a complement to embankments- could reduce the duration of flooding on the rare occasions when the embankments are overtopned. i/ 2/

4.106 Alternatives to the gravitv canal irrigation scheme nronosed in the report should be considered, particularly in view of the short dry season= Low-lift pumps and tubewells could provide attractive alternatives to gravity canals without the right-of-way problems created by the latter.

4.107 The project appears justified, at least for flood prevention And drAinnage. Addtiionna sttudIes shonild be made ennsidprtno altprnativpq suggested above, and economic studies should be prepared in accordance with current standardsa etahbllhed hy FPWAPfA in pnnnperanon with the East Pakistan Bank Group.

4.108 Terms of reference for revision of the 1966 feasibility study should be prepared an.d should cover, in additinn tn the alt-Prnativeas siug- gested above, adequate treatment of project organization and repayment of project c0sts; see paras 4.39, 4.40 and 4.60. lahnoldir.gs in the Kushiyara project area, which average 15 acres in size, are considerably

1/ Spillways would be needed to direct overflows and thus prevent erosion

oL the e-Wank.-LLIn.LB. LLCO 1m.LIL Ub LocadU CL thLeLat k ,lULtLet Wilere drainage structures (sluices) will also be required.

2/ The farmers' statement referred to above about the duration of overflow is conrfirmed by river-stage reoUrdu. nUe faCL LtiaL Ltle uratioU of overflow is much less than on the main rivers of the Province means tnat, if only 10-year protectLion is proviuea, tne auratLion or overrlow resulting from a larger flood might be only a few days while the volume of water entering the protected area might be sufficiently small so as not to cause serious damage. In reviewing this aspect of the project design, economic as well as hydrologic factors should be considered. - 61 -

above the norm for East Pakistan. Nevertheless, specific recommendations are needed on the magnitude and timing (probably on a progressive basis) of project repayment charges.

4.109 Once the terms of reference are prepared and considering the large amount of useful information in the 1966 report, it should be possible to prepare a revised feasibility study of adequate scope in about 12 months' time. If this can be accomDlished by June 1971 it should be possible, provided that financing can be arranged, to start construction by early 1972. Based unon nreliminarv information, costs of the proJect have been estimated at $13.5 million.

4.110 Khowai project: A feasibility report on the Khowai project dated Februarv 1968 wtas nrepnred fnr EPWAPDA by Associated Consulting Engineers, Dacca. Physical characteristics of the project divide it into twn distinct nrpAaF the Wnrthern Area, cnmnoRed nf 39,800 Acras which is flooded from late May to late October to a depth of 5 to 20 feet every year hv hackwnter flow fronm thp Dhanleswari Meghna, and the Southern Area encompassing a gross area of 40,200 acres with a net irrigable area of 330nnn acrre w nh is floodenA hw tho a ihnuniVnranai andti Su,tnan rive The Northern Area development has been designated by the consultants as Phase I and the Southern Area development Phase II.

The l^,.tra r<4-r 4S a rainfed stream with most of the drainage area in India. There are at least 30 to 40 pronounced flood peaks every year.*Nor-aal annual rainfall averages ab-bout 100 in.ches annr..ally, 86 er- cent of which falls in the six-month period April through September. As result. oc flooding from the Vhowai ri.ver anA t-o some extent from t"hLe Sutang and Karangi rivers, and the rainfall pattern of the area, crop proAuctilon is huampereA bLy flood anA drought.

XZ l U..Aa L5 LLJJ &A. L V ALLJI%.L " .L a. IC .L L J1 UL3W J.V .U 4.112~~~~IPrbl.J.-,s of flood.uA. preventlon ir,A -le.&LL 'lase I area 'belo-w elevation.s 20 to 25 feet can be relieved in a minor way by some channel improvements Sandemban.4 fn%.e.0s but most of thIe flooding is due toLthe backSwatCers. of thee Meghna and should be studied as a part of the flood prevention plans for other areas also f loo(dled,CW-g.eg,.lna. by thIUe However, fIloo protection can be achieved by embankments in the Southern Area designated by the consultants as Phase TI, an' such a plar 'as bueen recouulen,deU.-

4.113 .,.~~~efl'oodu protect',on scheu.e 'Lor Phase II -wvu invlv corsA. tion of embankments along both banks of the Khowai river, a river reach of

LU16.5 11UIALesu.ileL, eII r.LgLtr&h;J.h UanL'SUar. of. LLLthe 'utan.g1U~LL riverLLA.L 'or 26dU.,U milesULJ~ an'U -.LLAhe leftLL bank of the Karangi river for 17.5 miles and a diversion sluice and channLel from t.he Khowai. rear Chn-1ighat to the Suitang ri-ver. TIh UL0uULtants propose protection against floods of a frequency of around 200 years. It is questionable wnetner thLs mucn protection can be justified; a review of the plans to offer a lesser degree of protection - perhaps for a 25-year frequency flood - is therefore recommended. - 62 -

4.114 Provision of water supply for irrigation may prove to be very valuanhlp in thp Sout-hefrn Arepa during thed vy seasr The co-nultants planned to provide water for irrigation from a conventional gravity canal 4 s.ystem. Thev alan nrnnnopd on hibild a harrage on tho Vhw.a4 r ver anA divert water into two main canals totalling 21 miles in length. Forty m lesa of latearals tnu.- be required to serve all the land *.der this scheme.

.1!1 ~~ Graivlty -arl systes- require considerable right=of=way and cause significant impact upon the densely populated areas of East Pakistan such as the Khsowa area. The canals of recessity Must be located on the h'her ground where villages are usually built and as a result right-of-way costs are hi4gh, dAisplAaced people are very distb -nd the proces ofU btU 'rLI±g land is time-consuming and frequently delays project implementation in- efin4 + . *1.C altLeraL4VC L L-eCrllp W4-CL 'or rLLgatonu .nIILto urain- ageways and khals from which it can be lifted to the lands by low-lift pumpsa has many advantages and should be cons idered for the Khowai project. A second alternative that should be examined would be to use the borrow area alUJ4L;C11L LU Ltie outLer e-iubankmLent as the maiLn canallIUCations s-o tnat less right-of-way would be required.

4.116 The Phase II project seems economically justified for flood prever.tion andU iLrriLgatiLon. AddiUULtlional stuLies should beu- uae consiiering the alternatives suggested above, and economic studies should be prepared of adequate standard (see para 4.39). Ter-ms of reference for revision of the 1968 feasibility study should be prepared and should cover, in addition to the alternat'Lves suggesteu above, adequate treatment of project organization and repayment of project costs (see paras 4.40 and 4.60).

4.117 Once the terms of reference are prepared, and considering the large amount of useful information in the 1968 report, it shouid be possible to prepare a feasibility report in about twelve months' time. If this can be accomplished by june 1971 and if rinancing can be arranged it should be possible to start construction by the end of 1972. Based upon preliminary information including the i968 report, the project is estimated to cost $11 million. It would appear practical to employ the same consulting firm on Khowai as will be used on Upper Kushiyara.

4.118 Sangu multi-purpose project: A report on this project dated January 31, 1965, was prepared by Sandwell & Company Ltd., under the aus- pices of the External Aid Otfice, Government of Canada. The project is located near Chittagong. The irrigation aspects are presented in a separate report dated June 1964 prepared by Associated Consulting Engineers (ACE Ltd. of Karachi and Dacca, Pakistan). The Sandwell report contains reviews of portions of the ACE report.

4.119 The main features of the project are a 190-foot high storage dam, a diversion dam (barrage) at Dohazari and a gravity irrigation system covering a command area of 100,000 acres. The main dam would include a power house with three 27,500 kw units. Average annual generation would - 63 -

be 229 million kilowatt-hours at 33 percent plant factor. Tne benefit-cosi: ratio for the hydro features was computed at 2.11; for the irrigation features it would be 3.79 for an "optimum program" and 1.35 for a "'practical program"; for flood control it would be 2.43. 1/

4.120 The Sandwell report, states that more sedimentation data are needed to estimate the life expectancy of the reservoir.

4.121 Topographically, the irrigated area is low-lying; 80 percent of it is between elevations of 4 and 12 feet above sea level. It is of semi- bowl like shape with a depression in the center along the Chand Khali. The ACE report states that "this makes the task of canal alighment difficult and the system costly". Also there are "innumerable stream crossings".

4.122 The average cultivated holding is only 1.9 acres and with an average of 15.7 parcels per holding.

4.123 Rainfall in the area is relatively high. The average annual rainfall is 108 inches while the average October rainfall is 7.1 inches. Since these amounts of rainfall, particularly in October, could be ample for transplanted aman (provided proper cultural practices, e.g. diking of plots, are adopted by the farmers), relatively large benefits might be obtained from a first-stage project limited largely to flood control EIs described below.

4.124 From the map of the irrigable area, it appears that a considerable mileage of coastal embankments have already been constructed which may give some protection against tidal inundations. Maximum high tide conincident with high flood occurs in July and reaches 24.49 feet with an average maximum of 14.9. Average high tide during the low-flow months is between 4.3 and 6.1.

4.125 Considering the small farm size, the low elevations and the numerous natural channels in the irrigable area, an alternative project based on use of the natural channels as irrigation distributaries (supplemented by low-lift pumps) and, to the extent feasible, as navigation waterways, needs serious consideration.

4.126 A phased development of the irrigable area should be considered taking into account the extent of development that is possible preceding construction of the storage dam. For example, the project could be limited to flood control at first (to the extent possible without the storage reservoir) followed by supplemental kharif irrigation, which might be obtainable by operation of tidal sluices only.

!/ These ratios are considered illustrative only as the interest rate and foreign exchange rate, discounting procedures, etc., would need reconsideration= 4.127 The capacity of the hydro-electric installation should be recon- sidered for a possible lower plant factor and higher installed capacity taking into account changes in the power system that have occured since the report was prepared. The possibility of pumped storage for peaking should also be studied making use of the Dohazari barrage for re-regulating storage.

4.128 Aspects of the project that are scarcely treated in the 1965 report but are vital to success of the project are those dealing with organization and project repayment. (Although thus far only EPWAPDA has been carrying out the planning of this project, other agencies should be directly involved, even in the planning stage). These matters need treatment as discussed above (paras 4.39, 4.40 and 4.60).

4.129 Further steps to advance this project should be considered on a reconnaissance basis before proceeding to detailed revision of the existing feasibility report. It has been proposed that a single consulting frim or group be engaged to carry out the studies for the nearby Karnafuli irrigation project as well as for the Muhuri, Little Feni and Comilla-Noakhali prolects. It might be possible to arrange for the same firm to carry out the reconnaissance studies for the Sangu project and for a subsequent detailed revision of the existing feasibility study.

4.130 This proiect would appear to present no serious technical problems and might provide an excellent opportunity for a hydro-peaking plant to fit in with the East Pakistan power grid but adequate costs and benefits will not be available until further studies are completed. Assuming that the feasibility report and final engineering could be completed by the end of 1973, construction could begin early in 1974. Based upon preliminary information; the cost of the proiect has been estimated at US$50 million.

New Multi-PurDose Proiects -- Feasibilitv Studies Not Available

4.131 Four nroiJcts; covering about 2.5 million gross acres (1-8 million acres net) appear to be viable projects from technical and economic aspects based uotn information available and when comnared with other pronects that are now at advanced stages of preparation (especially Chandpur and Dacca Southwest). (rinpl,ia4nna regarding technicnl anid economir soundness have been based upon knowledge of topographic, soils and hydrologic factors and of existing cropping patterns. However, feasibility studies have not been made and these must be completed before conclusions are made to proceed 4 Jith- cor.struction. These pr- ects are:

D-aca N;orth. Barisal

T-LL-le LenL.L Comilla-Noakhali

4.132 The projects listed above involve no serious technical constraints and would appear to fit well into long-range developrm.ent needs oL East Pakistan (see Chapter 5). It is concluded that feasibility studies should - 65 - be undertaken as early as possible followed by final engineering and construction. (In most cases, since the outlook for economic iustification 1s favorable, final engineering could proceed on the basis of an "interim report" which could precede the completion of the feasibility report.) How- ever, the aggregate area of the 4 projects is large, and stage development: should be considered where nractical. At this iuncture it annears likely that,the initial phases would not involve more than 0.6 million acres (net:) nut nf a total of 1.8 million (net); sea Table 2.

4=l33 The Dacca North pronert area is lornted in the Northennt region of East Pakistan between the Dacca-Joydebpur Sripur Road and the Banar- Lakhya River, nnrth nf the rityi nf !)acca in Dacca D4str4ct. The area es- tends from Dacca Demra Road in the south to Sripur Railroad in the north.

4.140 The gross project area is 218,000 acres and plans are proposed to irrigate 174,000 acres. ElevatIons in the pro4ect area range from 4 feet in the center to above 30 feet near Kapsia. Higher elevations occur d genrerally -alon the river or. the easte .. side of the -rn4--* an along the western boundary. The high area along the west is rapidly developing as ;m industrial area.

4.141 The East Pakis-an Bank rGroup has h.ad te o a draft of an agreement on study of the feasibility of the project between EPWA!LIDA a censortlum of th.ree consulting f4i 'DaPkstar. Te.hr.-noronsult Ltd., Sir M. McDonald and Partner and Hunting Technical Services Ltd. The - - 'the IId- agreement wouAl.agraLuaLtIJU.U ULLdifre Lct I LLV L.PLLUJ.LGLLconsuiltants in U theLLLC stdL.UUY ofUL P1Y5L%..physica 4 plar.LULLWCU - economics for full development of project land and water resources to achieve LoLd prevention,LrrLgatLon andu UraiLnage. Several alternatives are suggested to achieve flood control including protecting the project area frolm LdakIhya River f.Loodi.g buy eankments anU a bUack levee diu dUraLn along the border with the jungle to the north. Gravity sluices and pumping plants would be studied for drainage. The source or irrigation water wouid be the Lakhya River through a pumping plant on the right bank. Study would be -auadeboth of' a gravLty canal'A systeml adu of a system mauIui'.ng use ofL ex'Lst'A;1g drainage channels as irrigation distributaries with low-lift pumps to lift water to the LandUs.

4.142 Ine plan of development would include stages starting with partial flood protection with provision for increasing it, and the irrigation system developed in blocks to the extent practical. Gradual draw- down of flooding would be planned so that farmers would not be subject to suuuen anu urastic aujustments by complete elimination of flood water upon project completion.

4.143 As there is some relatively higher ground within the project area, which may not be flooded or may be subject to only shallow flooding, it might be useful to consider two additional alternatives. One would be a first-stage development limited to the high ground with water obtained from tubewells. The second would also consider tubewells for the high ground but as a final stage of the project. - 66 -

4.144 The intent of the agreement is to make economic analyses using modern techniques. The terms of reference do not specifically mention repayment of project costs and this item should also be adequately studied in detail; see paras 4.39 and 4.40.

4.145 It was found in the study of Chandpur project that sample areas for study of detail plans and costs for low-lift pump implementation yielded valuable data. It would seem that the same approach should be followed on this project (see paragraph 2.14).

4.146 This project would appear to present no significant technical problems but adequate costs and benefits will not be available until the proposed feasibility studies are completed. However, the similarity of the project area to the Dacca Southwest and Chandpur prolect areas leads to the conclusion that similar benefits will be obtained and that the economic feasibility of the project will prove adequate.

4.147 Repayment and organization and management asnects of the pronect must be studied fully. Organization and management should be explored thoroughly and governmental responsibilities determined as has been done for the Chandpur Project.

4.148 EPWAPDA and its general consultants should follow the feasibility work closely to ensure that it is adequate. If this is done. and arrnm- plished by mid-1972, it should be possible, if financing can be arranged, to begin construction of the proiect in 1973. Based on prelimfnarv information, the cost of the project is estimated at $40 million (total equivalent).

4.149 The Barisal area comprises 1,500,000 acres in the southeast portion of the Southwest Reeion (see naragranhs 1-12-l=S, Mnn 2 and Tahle 2). It is a zone of ample sweet water and is subject to relatively shallow flooding. 4 The soil and water resources of the region 4ndirate that it ic one nf rhb h potential and where major benefits can be achieved both in the shortrun and over a longer time nperiod at relatiuely low cost.

4 4.150 The Barsa4l regio%n ic onei,wherein the nst alla by Anc of about 10,000 low-lift pumps has been contemplated and in fact, the estimate nf wAtav sailnh4lItt, am,-v4haA In n,-a A nl 4nA4'.-.t-A tHo- 4hnt,_-..1A Ie ~~~ ~~~~~~~v~~~~~~~~~~~s ~~~------I r_ ------vs feasible to install 10,000 such pumps in the region. Actual installation by ADn up to now has been only abo-t 1,000n du to slack demand from farmers in the region. The cause of the low demand is somewhat complex. One reason is that the-elatively large-a. .dholding in thIe aa -ve -ise 'Lo consLuider- able tenant occupancy. Another reason is that farmers already obtain an aus crop ir. about one=third ofL the~ aLre WhiJ.ch%.L, alhLL1ught lUW-y).ieU.LLdg , WoUlUU be competitive with the T aman crop that would result from low-lift pumping.

4.151 According to the UNDP soil studies as confirmed by questioning several farmers, much of the presently grown aman crop is not transpianted until late September or October when flooding in the area has receded. The depth of flooaing nowhere exceeds two to three feet. rnus, riood prevention embankments would have to be only five to six feet high which is much less than (for example) Cnandpur or Dacca Southwest. - 67 -

4.152 It can thus be visualized that development of the Barisal region would gaveh three stzages: (a)Io 4v-lift- (t-o the_ ext e_nt, of e_r -av A.±%A LI L.&IC L. % ..- ta . p'-.p8sfJ &.o ---t .. v Maa & ability); (b) further extension of low-lift pumping which would involve fixed primLary ptuping stai.ons ar.dULpossibly deepenin.g of some of the naturall char.== nels (khals); (c) flood prevention embankments and appurtenantworks includ-- ing drainage sluices. It MCCUMi likely that drai.age wVuld not be needed as the area could be drained during the low part of the tidal cycle. In some portions of the regior, moreover, wlhere tLUal actLoL s favorable, primary pumping might also not be needed for irrigation, thus eliminating stage (b) except for Udeepering of tihe k-hals. NatLUUal lVeee exitLL loUUg s of the streams in the region; in such cases, embankments could be omitted, at least initially, and only draiLnage sluices (supplemented by earth plugs to close off the khal exits) installed.

4.153 In view of the very large size of the Barisal area, it will have to be developed in phases. Phase I would cover 300,000 acres gross and 200,000 net irrigable. Project preparation, final engineering and capital investments for a Phase I project are outlined in Table 1. The cost of the project preparation is estimated at $4.0 million total equivalent. If thifi can start by january 1, 1971, it should be completed by June 30, 1973. The initial work of project preparation will have to include a reconnaissance of the entire 1.5 million acres in the Barisal region so as to delimit the 300,000 acres in Phase I. Final engineering is estimated ascosting $5.0 million total equivalent. This would include the cost of final engineering for all three stages in Phase I. Stage (a) estimated to cost $6.0 million total equivalent would be for 2,000 low-lift pumps and some drainage sluices in an area of about 100,000 acres net. Stage (b) estimated to cost about $14.0 million total equivalent would be for 2,000 additional low-lift pumps and primary pumping stations to serve them. Stage (c) (cost $16.0 million. total equivalent) is for flood control and drainage for the entire 300,000 acres (gross) in Phase I.

4.154 EPWAPDA and its general consultants have already prepared detailed terms of reference for consultants that will be required for project preparation and final engineering. These terms of reference which have been reviewed by the East Pakistan Bank Group include adequate treatment of economics and project organization (see paras 4.39 and 4.40). Effective inter-agency cooperation will be needed to ensure success of this important project. The East Pakistan Bank Group in its work program will endeavor to give particular attention to the development of the Barisal region in view of its high potential which is obtainable at relatively low cost.

4.155 The proposed Southwest Region Water Study (see Chapter 5) would be carried out concurrently with the project described in the preceding paragraphs. Going ahead with the Barisal project development need not, however, await the completion of the water study which it is judged will take about five years. The reason is that the Barisal region is in a sweet-water zone fed largely by the Brahmaputra River. The Barisal project would moreover have little effect on the water balance of the critical (western) part of the southwest region which depends primarily upon flows of the Ganges River. 4.156 Little Feni Project: According to the IECO Master Plan report of _ h..JL______Zcn , n n - . . na, , _ ~.¶_.. _ _ . _ _i_…. December 17OQ, t'ne lJlLL±tt Frli prUJt:>L WoULU llVav a groUs area Ut 2J57U0U0 acres and a net area of 150,000 acres. The project is located in the South- east region (see para. 1.16). in 1965, a large regulator structure (drainage sluice) was completed by EPWAPDA near the mouth of the Little Feni river to prevent tidal inundation and saline water intrusion. This structure, a preliminary design for which was contained in the IECO report of October 1961 entitled !!Little Feni and Noakhaii Regulators", has been described above in connection with the Muhuri project (see para 4.76).

4.157 As observed from the air, the Little Feni river contains many meanders. Some channel improvement work (cut-offs and iow dikes) have been constructed during the past ten years by EPWAPDA. It appears that additional flood prevention works in the lower portion of the Little Feni river, just upstream of the existing regulator, might have good justification, particularly if planned in conjunction with a possibie scheme tor water transfer to the Muhuri river (see description above of the Muhuri project).

4.158 Irrigation of the Little Feni project area, like flood control, would probably have to proceed in stages. Much would depend upon further development of the possible scheme, just mentioned, involving transfer of water from the Dakatia river to the Muhuri river via the Little Feni river. In working out such a scheme, the long-range requirements for the Comilla- Noakhali project, which adjoins the Little Feni project area on the west would also have to be taken into account. The first step, therefore, in the study of the Little Feni project should be a study, on a reconnaissance basis, of the proposed Dakatia-Little Feni-Muhuri water transfer. As pointed out in the discussion of the Muhuri project (paras 4.79 and 4.81) a highly important benefit from such a scheme would be navigation. Such a study has in fact already been programmed as part of the Karnafuli-Muhuri engineering credit (see paras 4.52, 4.79 and 4.86).

4.159 The reconnaissance study of the water transfer is expected to be ready by about April 1971. A feasibility study for the first phase of the Little Feni project could be started about January 1, 1972 and, taking into that considerable information would already be available from the reconnaissance study, the feasibility study could be completed by the end of 1972. The final engineering could start July 1, 1972, that is, before the feasibility study has been completed. Using procedures as have been worked out for other projects, construction could proceed on the basis of outline drawings and it could start in December 1973. As indicated in Table I, a four- year construction period has been assumed. The capital investment required for a Phase I proiect covering a gross area of 140,000 acres and a net area of 90,000 acres, has been estimated at $30 million total equivalent.

4.160 The Comilla-Noakhali Project lion maluly itn the Dakat{A R1ver drainaae basin in the Southeast ragton of I'ast P'akintan anlicovern a gross area of 892,000 acres (see para. 1.16). The area is bordered on the East bv the Little Feni and Muhuri project areas and by the Chandpur project (now under construction) on the southwest. 4.161 Present agricultural practices in the Comilla-Noakhali project area are li.Led princiLpO.Aly by floodin, which ir.volves fresa as we'll as saline water. Cropping intensity is presently only 146 percent. Yields are Low and BJ lluul L' L8le pr'LL ciLpa'l crop occupyiLng about 6I8Ipercer,t oL the cultivated land. Judging from the recent studies at Chandpur and Dacca

Soutnwest where croppiiig P"L1Ut z i flooding conditons are:- imi-la-Ir, present yields at Comilla-Noakhali could be tripled through water control.

4.162 An IDA credit of $2,400,000 has been granted to the Government of Pakistan to finance revised feasibility stud'es and final engineering of the Karnafuli and Muhuri Projects and consultants have already been engaged by EPWAPDA for this work. Studies of water suppiy ror tne Munuri project will involve investigation of the Meghna River as a source. The flat terrain and the presence of numerous natural drainage channels makes water transfer possibilities promising. On studying the water supply for the Muhuri project, the consultants will conduct a reconnaissance at the Comilla-Noakhali area to determine generally its water needs and its relationship to phased developments of water transfer schemes to serve as wiell the Muhuri and the Little Feni projects. Subsequent to such a reconnaissance it will be possible to determine what part of the project should constitute a first-stage development and should be the subject of a detailed feasibility study. Phasing of the project would be a crucial matter in view of its large size. Financial and organizational constraints would also need careful consideration before a first-stage project could be selected.

4.163 In planning flood prevention, the existence of the embankments of the Chandpur project in the southwest part of the Comilla-Noakhali project area could be taken advantage of. The drainage divide on the east between the Comilla-Noakhali project area and that of the Little Feni River is indistinct. Topographic surveys and hydrologic studies would be needed to determine whether flood spills are of significance and if so what pre- ventive works may be justified.

4.164 Based upon a knowledge of the topographic, hydrologic, and agricultural aspects of the area, it would appear likely that a technically and economically sound project can evolve.

4.165 EPWAPDA and its general consultants should make use of reconnaissance findings for the project, and prepare appropriate terms of reference for feasibility studies. Table 1 indicates the possibility of a construction start December 1, 1973 following procedures as outlined above for the Little Feni project. Based on preliminary data, cost of Phase I of the project (gross area 220,000 acres, net area 160,000 acres) has been estimated at $50 million total equivalent. Project preparation and final engineering would cost $3 and $5 million (total equivalent) respectively.

4.166 Ganges-Kobadak Jessore Unit: A report on Phase II of the Jessore Unit of the Ganges-Kobadak project was prepared by Dr. Mubashir Hasan! consulting civil engineer, dated October 1966. There are three units of the - 70 -

Ganges-Kobadak project: Kushtia, Khulna and Jessore. The Jessore Unit is f..rer subdA4ivi4Aded 4.into six "hases of which this pr 4 a tha - secon or ~~~~~~~r -- - -_ - rJ_-V Phase II. The report proposes that the Nabaganga River downstream of the MguraPa *t4--r b- used a a source of water supply for easten.. areas thle Fatki River, the central area; and the Chitra River, the southern area. Six puping plan.ts are proposed anA a series of canals fror. the pLuping plants totalling 174 miles in length. Additional works are proposed to pro-vide so..e I Aflocontrol and Arainage. The gross area is 140,00' acres of which about 105,000 acres is said to be irrigable.

4.167 The report points out that hydrologic data are extremely limited and that char.ges 'I.U 'lowsL brougLhIt aLboutL bUecause of the exi'.stiLng Ganges- Kobadak works (see Chapter III) have made many historic records useless. Sinice water supply to the eA.LstL.1g U1-UILL.L. it is Uy LIU me!ansl sULVed as yet, the situation with respect to the Jessore and Khulna units remains obscure. Besides the Garges as a source of supply, another major source that should be considered is groundwater. A third source could be return flow from upstream units, particularly 'n view of the high wastage or water that now prevails (Chapter III). A major part of the return flow from the Kushtia unit will flow to the Jessore and L\1&u±Lna -.iLts. The ydurology of the whole Southwest region is rather complex and actions are needed (see Cnapter 5) that will enable a gradual uwderstandiLng of the problem. Con- currently, however, it should be possible to proceed with the development of portions of the area basea on groundwater. inen, as reilaDie information is generated, surface sources could be utilized as well.

4.168 Quite possibley portions of the area will require drainage improvements including channel deepening and pumping and these facilities will also affect the water balance and therefore the sequence in which the various portions of the area can be irrigated and drained.

4.169 Whether a system of irrigation distribution based on use of natural channels supplemented by low-lift pumps rather than the gravity canal system proposed would also need investigation. Table 1 indicates preparation of a new feasibility study for the Ganges-Kobadak-Jessore unit (Phase II) at an estimated cost of $2.0 million total equivalent. The table also indicates start of such a study June 1, 1973 with completion a year later. An earlier start of such a feasibility study does not appear prac- tiable in view of the need for previous progress on the Southwest Region Water Study (see Chapter V).

Manpower Requirements

4.170 Engineering: Manpower requirements to carry out the engineering for the projects listed in Table 1 (includes the special studies and basic date programs described in Chapter V) are likelv to present serious problems. A largo number of expatriato specialists would be nceded in the first three years 1970/71 to 1972/73. Eveni assuming that thte home-office forces of the consultants will make up part of the work force and that there will be a - 71 _

maximum participation by Pakistani consultants and Government forces, the expatriates required to reside lr. Eas t Paklstan, fr the EPWAPDA nrn4ertr alone, total about 125 in 1970-71 and would increase to about 225 in 1971- 72 an.d 1972=73. Presently abouit An oe natr-4nrt e.nQ1ai,1tnntQ are naanocla with EPWAPDA projects and this number would have to be retained; the total require.uents for thtese tw years is thua 285. Paklst-anl pro^fesiralo4r. sultant needs will also be high with a 1970-71 demand estimated at about 200 increasin.g to .about350 in 1071=72 -nA 1072=73 S pf 1 East Pakistan in absorbing these personnel needs, particularly in regard to

hae expa tAr atS _. Expeienc Af AindicateC _4S -.. 4 A-A- eS&S-- A O-.t..AClA-- 4- 1.-. S SlSC CA~~~~~~~taLL -.. R- - 5 - ^C- a X f - - ^=W16z1 ^^Jz^> of expatraites will create public relations problems which must be dealt w.Ltitl carefully and at an ear'Ly ulate. Housing neeAs .-.ust be met if recruAlt ment is to be successful. Approval of consultants by the Government must be pror.mpt and eilficient. Salary scaLes must be reaL t'Lc andu attract'Lve. Schooling and medical needs for families will have to be recognized. EquLp- ment or alA. Klinu's suc[n a auLomobiles, office machlitGnes aridU enginXeLing.LL fin-- struments must be on hand when needed. While an effort should be made to locate consuiltants' offices outsiue of Dacca, practLca.l conssiderati'ons reLat- ing to location of Pakistani key agencies, facilities and information indicate that a large portion of the expatriates must De absorbed in Dacca. In addition to the need for consultants, EPWAPDA will face a large manpower need to assimilate the expanded program. Other agencies such as tne Surv,ey of Pakistan will also require increased manpower. EPWAPDA will need engineering staff, and a sizable body of sub-professional personnei ror operation and maintenance as the program progresses although this need will develop more slowly than that which will develop with the planning and construction of projects. If recruitment of consultants is unsuccessful due to failure to meet all the essential requirements cited above, slippage in the program listed in Table 1 will inevitably occur. The starting and completion dates shown for project preparation, tinal engineering and construction for the various projects listed assume that all the actions described above will be carried out without delays. To speed up administrative actions concerned, such as approval of consultants' contracts and approval of individuals to be resident in East Pakistan, existing procedures need to be streamlined, which may involve granting EPWAPDA more autonomy in such matters. GOEP has indicated its willingness to consider the necessary administrative changes.

4.171 Agriculture: In Chapter 5 of Volume I, a discussion of agricultural manpower required for both irrigated and non-irrigated areas indicates a shortfall of both Union Agricultural Assistants and Thana Agricultural Officers. Present training facilities are inadequate and both shortfallEs will get progressively worse (reaching a peak by 1974/75) unless a special training project is undertaken without delay. Such a project has been described in Chapter 5 of Volume I.

Agricultural Production

4.172 Table 3 contains a projection of the incremental agricultural production obtainable from the water development projects broken down according to low-lift pumps, non-intensive tubewells (constructed by ADC), - 72 - intensive tubewells (constructed by EPWAPDA) and multi-purpose projects. The proiection is Riven for the years 1971/72 through 1980/81. and an additional projection is shown for the year 1985/86. The projects included in makinR the vroiection are the three existing projects described in Chapter 3 (GK-Kushtia, Coastal Embankments and Chandpur) and the short-range proJects described in this Chapter 4. As has been indicated in Table 2, portions of the later phases of the Coastal Embankments and of these short- range proiects have been included in the projection to 1985/86 even though the capital investments indicated in Table 1 cover only the initial phases of the Coastal Embankments and of the larger multipurpose proiects (Pabna. Little Feni, Barisal and Comilla-Noakhali). The areas irrigated and provided with flood nrotection may be summarized as follows (in thousands of acres) 1/:

Flood Irrigated Protection

Present (1969-70) 860 -S80 2/ 1975-76 3,280 3,480 1980-81 5.960 5.730 1985-86 8,070 6,700 Ul11t-imanq tep 10Q380 8;050

4.173 Th.e nasmptions *Red in mAking thp nroiections were as follows!

a. Th.e inrePmPntal area irri4gated by low-lift numns would be 300,000 acres by 1971/72 building up to 1.1 m411ion acres in 1976/77. Therenfte-r it would decrease to 0.5 million acres in 1979/80 and would then remain constantt. The decrease is to nllow for the fact that an area of approximately 1.0 million acres overlaps with the EPWAPDA multi-purpose pro4ect areas.

b. The incremental area irrigated by non-intensive tubewells would amount to 50,000 acres in 1971/72 increasing to 800,000 acres in 1978/79 (this assumes 1,I 0 AAA nt t-ublewe'lJls b-y that d4at"e). ThS.ereafter the area irrigated by these tubewells would decrease to about30,00 Aacre by 1985t86.The reason for the decrease is because the area served would progressively be taken over by the intensLve-type tbe-WelI deve'Lom nUe.

1/ Area presently irrigated of 860,000 acres (see footnotes 2 to 7 in Table 2) has been added to grand total in Table 2. Area protected against floods in Brahmaputra Right Bank project (footnote 6) has also been added.

2/ Includes 1,000,000 acres in Coastal Embankment area. - 73 -

c. The intensive tubewell projects would have dates of start of rnnstruction as in Table 1. No area irrigated until 2 years after start of construction; thereafter fu:ll area irrigatPd rearhed after vpyars. The incrpmentAl area irrigated would build up to 1.0 million acres by 1980/81 and to 2.2 milio-n scres by 1Q85/86=

d. The EPWAPDA .ulti-purpose projects Jnrl,ide tihe 12 "rnaw projects" in the short-range list and the 3 partially co..pleted existingprJcs(osa bnk.t,GK Kushtia and Chandpur -- see Chapter 3). Dates of start 0f construc4ouldben ul A e a. i. Table 1. No ara 4irri4gated until 3 years after start of construction; thereafter, Cul1 area irrigated reachned- from, I t-o 7 years later- depending on size of project. The incremental area irrigated would reach 3.0 .million acres 4n 1980/81 -ar. 4.2 million acres in 1985/86.

e. The areas "effectively irrigated" in all cases except th_at of the low=lift pumps equals percent of: the L.ILI J..U Lil .L LL' .L U. L S D..DA/ASiO InJ ~J J S.L& .il L&LC incremental area irrigated during the first year, 60 percent Lthle secondA year andA 90J percent hIe third year. (Total elapsed period to reach 90 percent would thus be years LUo tUbCWell prLUJLet DanU LUo UIULL-pULrpose projects it would be from 6 to 10 years depending on the size of' the proJect area.) Ln thie case of tLe low-lift pumps, the incremental area effectively irrigated would be 40 percent the first year, 80 percent the second, and 90 percent the third year. f. For the multi-purpose projects, incremental production in tons was assumed equal to the area effectively irrigated times 3.0 tons. The latter figure was obtained from the Chandpur II appraisal report (in Volume IV), Annex 4, Table 4, taking future production at 341,000 tons (224,000 tons of paddy rice, which is equivalent to 150,000 tons of clean rice, plus 19i,000 tons of short-season crops) and present production at 85,000 tons (42,000 tons of clean rice plus 43,000 tons of short- season crops). The difference is 256,000 tons for a cultivated area of 89,000 acres or 2.9 tons per acre. The irrigated area is however only 75,000 acres which would give 3.4 tons per acre. To be conservative, 3.0 tons per irrigated acre was used. For the intensive and non-intensive tubewells projects, these tonnages per acre were assumed 2.5 and 2.0 respectively. The latter figure was obtained from the appraisal report included in Volume 1V. A slightly higher tonnage was assumed for the intensive tubewell projects since these would be located within a concentrated georgraphical area. For the low-lift pumps the tonnage was assumed at 1.2 tons per acre since the benefit is primarily from boro rice only. - 74 -

Periodic Review of Projects

4.174 It is proper and indeed imperative to proceed with the water development nroiects "in the pipeline" as above outlined. These proiects all have high rates of economic return, involve no technical difficulties and can be integrated into a lone-range Province-wide plan of water development including flood control (see para 4.176). During the next few years efforts will he made by other agencies to nrenare nurelv agricultural-innut nroiects (improved seeds, fertilizers, pest control and credit). It is possible that some of these nroierts will have rates of return hiQher than the water development projects. Moreover some of these projects could be in competition for the same finanrial and/or mannower resources. On the other hand, as several of these possibilities are at this stage only speculative, there is no reason to delay proceeding with steps toward imnlementation of the water development projects as has been outlined above. At the same time, it must be realized that the overall cost of the water program comes to $932.1 mIl- lion total equivalent (see Table 1), and that to carry out a program of this size ad 4r --codace with the Ates indAirated in Table 1 would require an effort of unprecendented magnitude. The manpower constraint, has been dis- -ussed ahbove and it was indirated that this contraint alone, as its effect is better understood, might have to require some shift (hopefully not more than asyear or two) in the dates for start of construction of some of the projects in Table 1. In order to evaluate properly the combined effects of marnpower and fina.cial constraints and of the com.etition for these resources by other projects, the array of projects "in the pipeline" should be subject to periodic review on an annual basis. This should obviously apply to all projects and should be done in cooperation with the Planning Department, ~%EP, -4t.ose staff would h-ave t-o be strengthened for this purpose (see_ Volume I).

Short-Range Projects -- Summary

4.175 From Table 2 and 4.172, the short-range projects may be regrouped as follows (thousands of acres): - 75 -

Flood Irrigation Protection 1985-86 Ultimate 1985-86 Ultimate

Low-lift pumps - Existing 700 700 - - - Future 500 500 - -

1,200 1,200

Tubewells - Existine 110 110 - - - Future 2,S00 2,500 - -

2,610 2,610

Multipurpose projects - Existing 50 50 1,580 1,580 - Future 4;170 6,480 5-120 6-470

4-220 6,593 6,7t00 8,A05

Total - Existing 860 860 1,580 1,'i80 - Future 7,170 9,480 5,120 6,470

Grand Total 8,030 10,340 6,700 8,050

4.176 With respect to the tubewell area, it was assumed that this would 4ncrease by 1O9.8.5-86 t- 2,610,000ac re an icasof 2, 5090,0 00 acres over the existing area. Although the figure of 2,610,000 acres is less than that

- *-.S lt1f ,...__ (n,w, I 1~' _ A , ^ _n -k l. *. aA ^ ffd . AA^^F @' Vf LthC favorale zone fypaa 4..t6)/ a-d can e _-LIL_ J^.e e a a tLarge, .L-hAA1-uAs is obvious that further investigations, as described elsewhere in this be neeAeA before the figure can be confl-e.Wt rsett volu.,ie,vJ.Lua willWL.t A. IJ * uC~A LaUJ I~ *.1t tOl L' .ft.~&I~A l. LI CO.L L L. the multipurpose projects, there are a total of 18 involved: 6 existing (of -WI-ich. 3 are oupleteUd)% anA 12 r.ew or.es - in other wordUs 1.5 prOJe s on, which major works are in the future. With one exception, there are no 4 4 technicalconsta A ts4 pAreventa nm-.. A91 A-V-1.pM-t aC thesA 1 - -Jc ta ea amplen LCLIILL3.LC.L UVI;AISILO L 03.1110 W VI..LI L.LIlfl ab_vMI.-J V~4.IJjJLI._Z ~ .A .Ot.- - r._J v On technical data exist (topography, hydrology and soils) and there are no un- resolveA lechnlcal problems of consequence. The one exception. is the Coast- al Embankment project where full development (including water management) is clearly attainable only -in h0alf. tIe proJect area (see para 3L.11). FVull development of the remaining half will probably have to await results -- as the-y become available -- of a special study of the Southwest Region.; see Chapter 5. The ultimate irrigable area, with this adjustment, would be reuuceu fror, a abou 91I.-illi acres. Theu LL.L±mate area with flood protection would however remain at about 8.1 million acres.

4.177 As above outlined, the "short-range" projects cover 9.1 million acres of irrigation (of w,.lch 0.9 million existing) and 8.1 million acres or flood prevention (of which 1.6 million existing). Development of additional "long-range"t projects are descrived in Chapter 5. - 76 -

CHAPTER 5

NEW PROJECTS -- LONG-RANGE

5.01 For the development of additional water-control projects beyond those described in Chapters 3 and 4, technical constraints both with respect to unresolved technical problems and availability of basic data become crucial. To overcome these constraints, various "special studies' will be needed. However, before discussing these, it is useful to consider the two main areas of technical constraints -- the flood control aspects and the water balance aspects -- and to consider how these affect the four regions into which the Province can conveniently be subdivided.

Flood Control Aspects

5.02 As explained above, the 12 new and 5 existing multipurpose proiects cover a gross area of 8.1 million acres. The Coastal Embankment project will provide protection basically against ocean flooding. Protection against other types of flooding (spill from the rivers and from excess rainfall backed up by high river stages) would thus amount to about 5.1 million acres.

5.03 The average area flooded in the monsoon season in East Pakistan is from 8 to 10 million acres but 17 million acres are subiect to flooding from time to time. Thus even with 5.1 million acres protected, a flood problem of considerable magnitude would remain. The problem of Drotectine the remaining areas becomes moreover progressively more complex and involving cnnsidprahle risks- East Pakistan consists of a larve delta located at the confluence of the Ganges and Brahmaputra -- two of the world's largest rivprR- They carrv at maximum stagp higher flood flnw than most rivers in the world and their sediment load (2.4 billion tons annually) exceeR thnt nf anv other river Rvytem in thp wnrld Rpratiie of their enormous transporting power, the Ganges and Brahmaputra rivers have eov-.rienced ma 4 nr shifta of their ehbnnnln in recent rentu,riepa Attemptnts to contain these rivers by "double embankments" may be extremely hazardous because of the very real danger of overtopping and breaching of the containing embankments and the impracticability of quickly evacuating the den.sel, populated zones in the areasa nrotected b the embran,k,ments,. (With one minor exception, none of the 17 projects mentioned in the preceding paragraph would involve double embankments. The one exception 4i the Dacca-Southwest project, the upper part of which would include over a stretCG of bout084 -lesan embar.cmn.t paralleling the e x is t i ng Breh,anutw Right Bank Embankment. However such limited extent of double embanking ts not considered serious afld ingfact wille proviof cost.rucion-. e below to be observed carefully following completion of construction -- see below with regard to flood models and observations on prototypes.) The problem of flood control is further complicated by the lack of materials, principally stone and gravel, that are required in the construction of works for river training and embankment revetment (the need for such work becomes intensified once double embanking is embarked upon on a significant scale) and by the fact that little is known about the ultimate effect of upstream projects (in India) that involve embankments (and even some double embankments).

5.04 While elimination of flooding is clearly crucial to East Pakistan -- whether in human or economic terms -- it must nevertheless be remembered that the local farmers have over the centuries adjusted remarkably to the annual cycle of inundation. This is most evident with respect to the prevailing crops grown, especially long-stem broadcast aman (also called "floating rice") which, although low-yielding, not only tolerates flooding but actually depends on it. Higher yielding transplanted aman is also widely grown where flooding depths are shallow and rainfall is favorable. Water control, through elimination of deep flooding and provision of irrigation, will permit planting of higher yielding varieties but it must be realized that sudden changes may not always be acceptable to or manageable by the farmers. Three examples come to mind in this connection. One is the Coastal Embankment project; elimination of fresh-water flooding in some of the polders included in the project has forced consideration of "1water management" in the interior of the polders. Another is Chandpur, where it is proposed that flooding be eliminated gradually over a period of several years to allow farmers time to adjust to the new conditions. A third example is the Brahmaputra Right Bank area where farmers have largely adjusted to elimination of flooding but where it is evident that further major benefits are realizable through introduction of irrigation.

Water Balance Aspects

5.05 During the course of the next 3 to 5 years, provided that the program of projects, studies and basic data gathering as outlined herein proceeds, a growing body of information on the hydrology of East Pakistan will become available. With this information "water balance studies" wiLl become possible, first on a regional basis and ultimately on a province- wide basis. Such water balance studies will be essential to assist in establishing the priorities of further proiects to be developed in the future and in evaluating their effects on water supply requirements for various uses including irrigation, navigation and fish culture.

5.06 The water balance studies will also be useful in considering the relative priorities as between the development of groundwater on the one hand (nee descrintion of the US Geolotecal Survey-assisted study in Chapter 4) and major surface water diversions by means of major structures like the Ganges and Brahmanutra barrages on the other (see below under Special Studies). 5.07 Insofar as the 17 multipurpose projects are concerned, it is felt that allocation of water use during approximately the next decade or two will present no problem. The reasons are given in the following subparagraphs:

a. Of the 17 proiects, those that depend on the flow of the Brahmaputra River are as follows: 1/

Chandpur 75,000 acres (net) Dacca Southwest (Polders 1 and 4) 260,000 " " Muhuri 42,000 Pabna (2/3 of area assuming 1/3 from groundwater) 2 0 0 , 0 0 0 •I " " Belkuchi 70.000-,, Little Feni 150,000-"' Upper Kushiyara 40,000 Khowai 33,000 Gomitla-Noakhali 650_00O-' " Barisal 830,000-

Total 2,350,000 acres (net)

b. To the above should be added the area irrigated during the drlv season hb lnw-lift numns under the AMC nrogram The total area in that program may in a few years reach about 1,500,000 acres but a large nart of this would he included within the areas of many of the projects listed above while another part would nnt hge serued hv treams of the nrahmanutra svstem= Current information is inadequate to estimate how much area is in these two parts and how ich in the remainder; A a rough guess it would be some 500,000 acres.

c. In the future other projects such as the Faridpur Unit of the 4 'ov Ano,_Rr-flav4 l nrnn4prt lnnl the qiireaquwar nrniert (total not- area of these two projects about 800,000 acres) could be developed, possib!y without flood control. It can thus be experraA that in a time period of something like 15 to 20 years (allowing ti4- fo:. construction and attainment of full devel-opmennt) the total area requiring surface water from the Brahmaputra coulAd reach about. 3. mllion acres .

lt Dacca North with 126,000 acres (net) has been omitted because it would obtain its water supply from the Meghna River downstream of Bliairab Bazaar. The low flow of the lBrahmaputra is assumed equal to the sum of the flow of the Brahmaputra at Bahadurabad and the Meghlna at L3halrab Bazaar (see para. h). 2/ Ultimate development. - 79 -

d. All of the polder-type developments within this area would presumably be based on the system conceived for Chandpur. This is a "closed system" and differs funda- mentally from the Ganges-Kobadak type of system (large gravity canals) in that the return flow as well as waste water remains within the polder and does not leave the project area. Thus only "make-up" water needs to be supplied equal to crop water requirements plus losses for: (a) evaporation from the natural drainage channels used as irrigation distributaries (about 1 percent of the irrigated area) and (b) seepage at the periphery of the polder. 1/ The make-up water equals consumptive use plus losses less rainfall. Monthly crop water requirements would appear to be highest during April; however, February may be a more critical month as it is the month of lowest flow of the Brahmaputra River. e. In order of their water requirements ner acre, the nrintipal crops that it is expected will be grown during February and April are as follows:

Februarv Anril

Boro ric-e, sugarcane, Aus rice, 4ute wheat, vegetables, sugarcane, fruits nutlses oilseeds, fodder, fruits f. Since the water requirement of boro rice is particularly high, the nronortion that will he grown in the future is of great importance. In flood-prone areas, once flods have been prevented, the farmers can be expected to grow two rice crops consisting of aus followed by aman (both making maximum use of kharif rainfall). When.

No allowance is made for operating waste and/or seepane in canals or field channels since -- unlike a conventional gravity system -- these wastes will all return to the drainage channels. There should moreover be no operating wastes even if the low-lift pumps operate on]y (sayq)18 hours instead of 24 hours a day duiring the time of peak cleamnd. This is because the fixed, primary pumps can operate 24 hours a day and there is mnipe snace in the natural drainage system to store even up to 12 hours of primary pumping. An allowance for deep percolal:ion loss is also considered unnecessary since under full year-round ir,ri- gation the water table can be expected to rise (IECO report page B VI]: - 32). An allowance of 0.5" per month should be adequate for the two losses mentioned. - 80 -

this is done, boro rice will gradually be eliminated and replaced by a short-season rabi crop. If for present purposes (which is to estimate water requirements during the next one or two decades) it is nevertheless assumed that boro rice will increase in importance in the medium term, a conservative cropping pattern (again from the point of view of water requirements) might be as follows:

February April

Boro rice 30 - Aus rice - 68 Jute - 10 Sugarcane , 2 2 Other crops-' 63 - Fallow 5 20

Total 100% 100%

R. Consumptive use factors for East Pakistan is a matter requiring research. Approximate values can however be estimated from the Blaney-Criddle fo--rmula as was done in the IECO Master Plan report of 1964 (Supplement B, Table B VII-17). Using data for Dacca as reoresentinR average conditions gives weighted average requirements as follows (in inches):

February April

Rainfall Mean year 1.0 5.8 75% dry year 0.5 3.0

Consumotive use-/ 4.1 6.3

Loss-/ 0.5 0.5

CrFn rFeuirpmunt Mean year 3.6 1.0 7sz drv yeAr 4.1 _ 8

1/ Pulses, oilseeds, wheat, vegetables, fruit and fodder.

2/ K factors as in Table cited: K = 0.2 used for fallow area.

3/ See para. d. - 81a

The number of acres irrigable per cusec then become:

February April

Mean year 180 700 75% dry year 160 190

h. Use of the 75 percent dry year (1 year in 4) would seem reasonable if assumed to coincide with the low-flow year of record as regards the Brahmaputra. Using an average figure of 175 acres per cusec for both February and April would give a water supply requirement of 20,000 cusecs. In comparison, the minimum monthly flow of the Brahmaputra River 1/ in February is 133,000 cusecs and in April it is 208,000 cusecs. Thus the diversion requirement would be about one-sixth of the minimum February flow and one-eighth of the minimum April flow.

i. In para. f. it was mentioned that boro rice could eventually be replaced by aus. This would result in a higher percentage of fallow in February as well as a hiRher vercentage of aus in April. If the fallow percentage for April is reduced to 5 percent and the aus area increased accordingly the diversion requirement for April would increase from about one-eighth to one-sixth. However. diverting one-sixth of the April flow (208,000 cusecs) would be far less serious than diverting one- sixth of the Februarv flow (133.000 cusecs).

1. The fore2oin2 analysis. although admittedly somewhat crude and subject to further refinement, indicates clearly nevertheless that the low flow of the Brahmaputra River is much more than ample to supply an aggregate area of 3.6 million acres. On the other hand. the diversions of the amounts indicated are bound to have some unfavorable effects with respect to navigation; saline-water intrusion and increased pumping lifts affecting the primary pumping 8tations of the nroiee.t mentioned. A pnauihle fartnr tending to mitigate these effects would be the development nf grouniidwater in unatrenTr Areans n in the Northwest region. where advantageous use of storage within the underground aquIfer might reasult in an incresne in the Au4ilablelow flow of the Brahmaputra River. In any case there does not appear to be any prospective better use for the amounts dlverted than to devote them to irrigation of the 3.6 m.ill4nnt-r ares noa lsesvc4led

Adding flow of the Brahmaputra at Bahadurabad and that of the neghna at Bhairab Bazaar (IECO Report, Voi. i, p. 80). Floating Pumping Stations

5.08 Floating pumping stations have been suggested to provide a means of diverting water from rivers and waterways. Such pumping stations can be designed for a wide range of conditions and sizes but the complexity of design and construction, physical adaptability, cost and economics must be considered in each case, and compared with alternatives.

5.09 Detailed consideration is already being given to large floating pumping stations to serve the Sureswar and Faridpur projects. These are located on the right bank of the Ganges-Padma river on the opposite bank from the Dacca Southwest project (see Map 1). The floating pumping stations would be located in the Ganges-Padma and connected with the shore by large pipelines floated on pontoons. The General Consultants to EPWAPDA are also studying floating pumping plants designed to fit a range of East Pakistan conditions.

5.10 In comparing costs of floating pumping stations with conventional intakes it is most important to fully determine operation and maintenance costs as well as capital costs. Geomorphologic and sedi- mentary conditions are factors that must be fully understood in apprais- ing the alternatives.

5.11 Floating pumping stations of large capacity involve many complexities of engineering design. On the larger rivers, transmission of water from the pump site to the land will involve a long, high- capacity conduit properly anchored, and/or floated on pontoons. Anchor- ing the pumping station and the conduit must be compatible with field conditions.

5.12 With respect to the 12 new multi-purpose projects for the short- range, there appears to be little likelihood for the applicability of floating pumping stations since these projects are all located adiacent to river or stream channels of relatively stable characteristics. Pumping stations of conventional design (these must of course be carefully sited and designed to minimize sediment problems) therefore appear to be better suited. 1/ For the longer-range proiects such as Sureswar and Faridpur. the floating pumping station alternative appears to be worth considering.

1/ This is borne out by experience with the Dacca-Demra pumping station is located or a relatively stable river and has suffered lttle or no difficulty from sediment problems as contrasted with the Ganges- and ahn station wiicr ios (scated on the highly u-stable3 Canges River and lhas suffered seriously (see Clhapter 3). - 83 -

5J.13 Terms of reLerence fILor feasibility stuUies of thLe long-rarige projects where irrigation is involved should therefore include study of floating pumping stations where conuditiLons are such tnat this alternative appears to have practical application. The General Consultants to EPWAPDA should develop some practical guidelines for determining wnen and under what conditions this alternative should be included in terms of referencesu for consultants making feasibility studies.

Regions

5.14 The four regions into which the province can be conveniently subdivided have been described in Chapter 1. The division into four regions is feasible because the hydrologic characteristics of each region is generally distinct. The Southwest region is discussed below under Special Studies.

5.15 For the southeast region practically no special studies will be needed. This is basically because all of the projects in the region are on rivers that are independent of the Brahmaputra-Ganges system. Flood prevention will generally present no technological difficulty (except for the relatively small -- 47,000 acres -- Meghna-Dhonagoda project) and water supply can be obtained mainly from the relatively stable natural channel of the Dakatia River which connects with the Meghna.

5.16 In the northwest region, except for the Belkuchi and Pabna project areas, groundwater rather than surface water may well prove to be a more desirable source of water supply. Of the 4 tubewell projects mentioned in para. 4.25, 3 are in the Northwest region. Further tubewell projects in the region can be expected to be identified. On the other hand, integration of surface and groundwater supplies may well prove advantageous in the long run; a gradual generation of knowledge on this matter would emerge from the proposed special study "Brahmaputra- Ganges-Meghna Complex," described below.

5.17 With respect to the northeast region, projects that are believed should go ahead as rapidly as possible are Dacca Southwest and Dacca North and several relatively small tributary projects such as Upper Kushiyara and Khowai. A tubewell development in the North Mymensingh area also appears feasible and is one of the four projects mentioned in para. 4.25. The so-called Brahmaputra Left Bank Embankment project is located in the Northeast region. It however presents serious technological difficulty and as a preliminary step flood models are needed as described below.

5.18 Most of the remainder of the northeast region is subject to deep flooding from the Meghna. Working out a practicable means of mitigating the effects of such flooding needs a great deal of study. A possible - 84 -

approach is that of submersible embankments as has been suggested for the Meghna Dhonagoda project in the southeast region. Alternatively, it could be accepted that the area of deep flooding be set aside for use as a storage reservoir with resulting benefits to flood control as well as low-flow augmentation. Such a scheme, which would be of rather grandiose scale and involving difficult technical problems, could be examined in a preliminary manner as part of the special study "Brahmaputra-Ganges- Meghna Complex."

Special Studies

5.19 The special studies needed for the preparation of long-range new projects are as follows:

Ganges Barrage and Southwest Region Water Study Model of Southwest Region Flood Models Brahmaputra-Ganges-Meghna Complex (Preliminary Planning Study)

(In addition, various comprehensive studies, which are described in Volume I. would be needed including (a) a study of financing and repayment of projects, (b) a manpower study, (c) a study of infrastructure, (d) svstems analvses. and (e) an overall svnthesis. Certain of the comprehensive studies, such as (a) and (b), are crucial for the short- range as well as the long-range prolects.)

S 20 GanBes barrage and southwest region water study: By far the largest investments to date have been made in the southwest region. These amount to shout IUSS200 million for the Coastal Emhankment nroiect anei $60 million for the GK-Kushtia project. From the hydrologic point of view, the sou,thwest region is by far the mnot conmplex (see naras 1=12-1=1 This complexity results from the following physical characteristics:

a. There is a large tidal range -- of the order of 10 feet; tidal csurges from typhoons may reach elevations 25 feet above mean sea level.

b. There are large tidal estuaries whichi vary in width from one to severnaiVl mls GP llre of theea estuaries would provide benefits through elimination of sanlimne-water penetrnation and through flood prevention, but would entail major engineering works re- qulring careful study. The effect on fish ecology would also need to be evaluated. - 85 -

c. Flows enter the area from the Ganges and Brahmaputra Rivers. The latter is much greater in volume than the former. When the Farakka Barrage in India is completed, the flow from the Ganges River may be reduced which would increase the saline-water penetration especially in the western part of the region. 1/

d. There is evidence of a large groundwater potential particularly in the western and higher part of the region but actual availability of groundwater needs quantification. The USGS groundwater study program would assist in this effort.

e. Thp region iR interlaced with natural rhanneln. These vary greatly in width, depth and state of enlargement or devay Many c-ontrol struitutirpes wotltid hp neepded to enhance drainage, water distribution, capture of return flnow from tinnrpam arpas; navioAtioln and fishpries development.

f. EPWAPDA has the intention to proceed with the detailed structural design of n ,nior hArrage across the Ganges River. While such detailed design may be useful, it will be necessaryeto stud- the barra-e from a furnc4tioa1 as 'ell as a structural viewpoint. To determine the functional requirements of such a barrage will require consideration of the above five features followed by an analysis of the re-ional water balance.

5.21 A period of a year should be spent in reconnaissance of the southwest region as soon as this can be organized, possibly by January 1971. 2/ Te reonnaissance sh..ould le folowA by. aa. p...aly t year detailed technical study, making four years in all. The purpose of the reconnaissance wonuld be to outline and get agreement on the terms of reference of the detailed study. The services of an agency with experience on large tidal deltas, should ae sought to conduct both the reconnaissance and the detailed study. The cost of the 3-year study is eS-i.MateAast $5.dt. milon (total equilvalent). Thi Ascnlee 55L.UaL a' -_ .J-j a \~J~'LI-. ~.J VaJL %/ 2_It-4&.I ILIJ0 LO Lo LOLUL~ minimal and in the long run could be exceeded by a substantial amount.

1/ The present average flow during the three low-flow months (March, April and May) is about 66,000 cusecs. The minimum 3-month flow in 35 years of record (1934-68) was 55,000 cusecs in 1953. One year in 7 can be expected to have a flow of about 60,000 cusecs. (From "Ganes Barrage Project", Associated Consulting Engineers, Dacca, 1969.)

2/ Table 1 shows a starting date of January 1, 1972. The reconnaissance would start a year earlier as part of the Comprehensive Study (Chapiter 8, Volume I). - 86 -

Uncertainty regarding the cost estimate stems largely from current lack of knowledge regarding the extent ot surveys, measurements and subsurface explorations required. By the end of the reconnaissance it should be possible to firm up the cost of the entire study. Except for the Coastal Embankments (Chapter 3) and Barisal (Chapter 4) projects it seems improbable that initial projects in the Southwest region could be formulated prior to about 1974, assuming that the detailed water study gets underway in 1972.

5.22 Model of southwest region: An indisoensRAhI analvt-irni tonl in carrying out the southwest region water study would be the setting up of a mathematical model. Such a model including it onperat-ion -woulst about $2 million (total equivalent). Based on reconnaissance work that would have been carried out for the Southwent region nQ decrlb4hd in the foregoing paragraph, work on the model could begin July 1, 1971 as indiraterd in Table 1.

5.2i Flood models: Desnite the terhniral difficulties described above, it is recognized that moving forward with a long-range program of flood Drotection should be expedited. 1/ It should also he recognized that since the problem is urgent there may not be sufficient time to rollert all of the nhvsical data normally renuired in rhe s-tudy of s.u. complex technical problems. It may therefore be necessary to proceed with some works that would nrovide nrototypes that can be observed in order to assist further planning and action. An example of such a pro- totvne that might not involve undue risk is that to he provided by apart of the Dacca Southwest project. For projects such as the Brahmaputra Left Rank heowever (narar 517); the risk involved in const-riuct-ion of the pt type would be excessive. To assist in solving the technical problems for evalsinting the risQks for suicih a nro ert twmo helpful steps would be the use of flood models and the establishment of a Flood Consulting Board (see para. 5.30). The flood mndeil wonulfd he of rw0 tt,re: physical and mathematical. The physical model would consist of a scale model of a repre- sentative reach of the Brahmaputra River, proba-bly in its upper pri between Bahadurabad and Sirajganj. (Other reaches of the river might be modelled a'-so but atl a later stage.) Teeare serixous techiitica'l diffl'- culties involved in the construction and operation of a physical model

1/ This has been pointed out by various leading experts. See: United Nations Water Control Mission, Water and Power Development in East Pakistan (the Krug Mission report), United Nations Technical Assistance Programme, New York, 1959.

J. Th. Thijsse, Report onHydWdrology of East Pakistan, May/October 1964.

J. Th. Thilsse, Additional Report on liydro ogyof East Pakistan, Harch/Aprri 19065. - 87 -

since, Lor exauIple, iLt As ULLLM. asWLL y etL WLLhLet&hL Lthe alluV.iLUm that COr.ELLC.Ati tute the bed and banks of the river can satisfactorily be reproduced in a

moue anu Lte LU tiCLLC vvc.y C raL C xSU.iI3LIL uIuvenutseLL %ueu .L0aLU and suspended load) of the river. Nevertheless because of its crucial importance, it is cnsiduereu enat an attempt to set up a physical modul is well worth the effort. (See also below regarding a physical model oi, a proposed barrage across the Brahmaputra.)

5.24 A mathematical model would also present formidable technical problems but here again, in view of the possible beneficial results and the very major works that mignt uitimately be considered ror constructiDn, the indicated investments shown in Table 1 for the physical and mathematical models are considered justified. Assuming that preparatory work will have been performed by the Flood Consulting Board described below (para. 5.30), it should be possible to start work on the flood models July 1, 1971 as indicated in Table 1. The proposed physical and mathematical models are estimated to cost $2.5 million.

5.25 Brahmapucra-Ganges-Meghna Complex: The construction of a complex involving a barrage across the Brahmaputra River near the international border with provision for diversion of flows to the Meghna Basin and to the Northwest and Southwest regions has been proposed from time to time. The purposes envisaged tor this complex are: to provide partial relief of flooding along the Brahmaputra by diversion through the Old BrahmaputraL River into the Sylhet depression of the Meghna Basin (see para 5.18); to divert irrigation supplies to serve water-deficient areas of the North- west and Southwest regions; and to generate hydro-electric power which could conserve scarce fuel resources and also provide energy which may be required for pump drainage, particularly in the Northeast region.

5.26 Development of an adequate safe design for a barrage across the Brahmaputra River would involve complex problems, particularly with respect to hydrologic and geotechnical features. The design would have to incorporate extensive training works to ensure stable conditions so that the river would not change course. An alternative massive groundwater development (paira 5.36) would have to be studied concurrently.

5.27 From a structural point of view, a barrage across the Brahmaputra would involve technical problems of great magnitude. The barrage would have to pass a very large flood flow (of perhaps 5 million cusecs) and it would have to be constructed on riverbed sediments of unknown depths -- perhaps several hundred feet. The prevention of seepage under such a barrage would require expensive cutoff construction or in lieu of this an extensive blanket probably of clay. To protect the blanket against wave action, rock or concrete blocks would be needed. Rocks and aggregate for concrete are either unobtainable or very expensive. A physical model of such a barrage would be needed to assist in determining the feasibility of any design proposed and also to develop a practical method for excluding sediments from an intake to a link canal on the right bank (required to transfer water to the Southwest region) and from an opening on the left bank (to divert excess flood flows to the channel of the Old Brahmaputra River as a flood control measure). - 88 -

5.2LO in adaition to th-e -model nvestigations, it wouId eventually be necessary, in order to make even rough construction cost estimates, to conduct extensive foundation and material investigations.

5.29 In order to define more specifically the various elements and functions and relate these to projected requirements, a preliminary planning study will precede the detailed investLigations aescriDea in paras 5.27 ana 5.28. This preliminary phase would have an estimated cost of $5.0 million total equivalent (see Table 1) and would be carried out concurrently with the related studies described in paras 5.20-5.24 as well as with the Comprehensive Study (Voi. I, Chapter 8j). All of these studies would be guided by a Flood Consulting Group as described in the next section.

Flood Consulting Group

5.30 The scope and cost of the various investigations described above, particularly for the flood models and the Brahmaputra-Ganges- Meghna Complex, are so vast that they would have to be planned with considerable care. For this reason, it would be desirable to set up a Flood Consulting Group that would direct and review the investigations of the various related components. The Government of Pakistan recently requested the Bank to send a team of experts ("Flood Commission") to assist in the study of flood problems. It was felt that since this problem exceeds in complexity any situation of like character anywhere, world-renowned experts of the highest calibre would be required in the various technical fields involved including hydrology, engineering, sedimentology, geomorphology, and geography. While the need for such a group is still felt, it is now believed that its scope should be broader and not limited to the flood problem alone.

5.31 The setting up of a Flood Consulting Group is therefore proposed. It would consider not only the overall flood problem but also the question of flood models and how best to proceed with study of the Brahmaputra- Ganges-Meghna Complex. It would probably also be desirable for the Group to give some consideration to the effects of upstream developments in India (see paras. 5.03 and 5.20). Such a Group might consist of 5 or 6 experts that would include the fields already mentioned but in addition would include a broad-gauged economist with experience in water resource development. Suggested terms of reference and costs for such a Consulting Group, which would be needed on a part-time (about 1/3) basis during an initial 4-year period, are discussed in Chapter 8 of Volume I. - 89 -

Basic Data

5.32 Much basic data are already available in East Pakistan on hydrology, topography, soils, agriculture and economics. 1/ These data are generally adequate in the short run for the planning or water projects, including those already underway which might start during the next few years (includes those for which capital investments are shown in Table 1). ltowever, for further planning, including the various special studies described above, additional data will be needed in the near future. Reviews of needs for additional data have been made and are being continued by the General Consultants to EPWAPDA. Harvard University is also working on basic data requirements and inputs and is exploring establishment of a basic data bank tor storage and retrieval of this information (see chapter on Comprehensive Study in Volume I). Need for additional basic data for long-range planning has long been recognized. A Bank mission of November 1967 observed that additional data are required particularly for study of flood control when double embankments are considered and that groundwater data are essential for study of an expanded water development program.

5.33 An expanded program of water resource development including agricultural implementation will also require an intensified program of aerial photography and mapping. Flood control considerations will be dependent upon additional photography and mapping also and will require basic data on sedimentology and morphology. Construction of river training works and barrages involve problems of finding suitable construction material (rock and aggregate for concrete are extremely scarce in East Pakistan). There is a void of information, laboratory facilities, and knowledge on geotechnical aspects of soils and bed materials needed for design of major structures. East Pakistan is situated on an active tectonic zone and data are lacking on seismology. Large hydraulic structures should be designed to withstand earthquake forces by making use of appropriate local data.

5.34 Data needs in surface water hydrology, groundwater hydrology, geotechnical information, construction materials, seismology, geomorphology, and sedimentology are discussed in the following paragraphs.

5.35 Surface water hydrology: Much data have been gathered by various groups and agencies and FAO has given valuable assistance with their program involving the training of Pakistani personnel to make gagings of streams. However much remains to be done in reviewing historical records to eliminate unreliable data, correct errors, establish procedures for recording records and retrieval of records when needed for use in hydrologic studies. There also is a need for additional training of Pakistani

1/ The presently available data were obtained through programs of EPWAPDA, ADG, p,An and the Survey of Pakistan (witn UNDP support) and through feasibility studies of specific projects, described in Chapter 4. - 90 - personnel in gaging large rivers and more modern equipment will be needed for this work. Accurate hydrologic measurements are also dependent upon re-calibration of measuring instruments at regular intervals. There is at present no operational facility in East Pakistan for accomplishing this obJective. Funds must be made available to establish this facility and trainina nrovided for its use. The additional nrogram for surfare water hydrology must be undertaken now to supply data needed for long- range flood nrotertion plans and to nermit comorehens{ve planning as in the Southwest Region and for water-balance studies. The estimated cost over a 4-vpar nprind- .Qwaq indirartpt in Tahlp 1 2 nf VJnl,ma T 4S $1.825 million total equivalent of which $1.5 million in foreign exchange.

5.36 Groundwater hydrology: Groundwater development in East Pakistan haa pr e adendi11 anon No intanclfi d (aAs rhaptair Li. Dataare adequate to proceed with the ADC tubewell programs (with dispersed well a,4 A.-A . A n-nt aF f nr fl raf- nh naee nf e-f-ha PPWATA ube-wePW projects. As more concentrated developments proceed, more comprehensive Aata will be neeAed both for project planning and for water-balance stuiles (see paras. 5.05 and 5.06). With progressive growth of hydrogeologic infnr-mat4inn msci. tuiihpwell dulnnmnts (lnrog canpa-it-y wells of over 4 cusecs and over 500 feet deep) will ultimately require consideration, possibly as an -alten-ativeto o-am,a 4or bsrra8e (paras. 4-10 and 5.26).

5.37 HyoA--g1e-.o,g s 4t . wil-l,,A-i11 4nca the f-d-rminti4on of aquifer characteristics (such as area, depth, material, specific yield and rechnarge), -well spacin and nt-her fctorsva peavrtinent ti,-4 tr.s- i groundwater development. Such studies have been programmmed by U.S. Geological Survey (financed by USAT D) to start in mid-1970 and to be completed in five years. The USGS studies will provide data for spe- ciLLfLc proJects but, as he proJects proceed, the UTSre studles will also make use of data generated by the specific projects.

5.38 Aerial mapping: Work should begin as soon as possible on a quality aerila.L map-pipng sur-vey ofL Eolst P'-istan. TLILs survey will be ar essential source of basic data for flood protection planning and the more comprehensive regional and multiple purpose de-velopments now being proposed. Information from this survey will be especially vital to the proper examination of double embankment flooa control projects, regional studies such as that proposed for the Southwest Region and comprehensive study programs such as the Brahmaputra-Ganges-Meghna Complex. intensified agriculture programs will also involve use of aerial photography.

5.39 Pakistani personnel should be trained and equipped to make necessary resurveys as needed in the future. As an essential tool in the study of embankments and river-training works, photographs of the main rivers both during the flood season and during the dry season will be needed so as to obtain information regarding the shifting of the riverbanks and streambeds. The survey will also provide basic data on topography, extent of flooding, location and width of waterways and land capability information. - 91 -

5.40 While most of East Pakistan is mapped with one-foot contours at a scale of 4 inches to the mile, and some to large scales, the maps vary greatly in quality. The nature of the proposed projects is such that accurate locations and cross-sections (including elevations) of drainageways, and land ownership are of great importance. Topographic details and property boundaries have to be determined exactly for these reasons. It has been found necessary to map, by ground methods, project areas at a scale of 16 inches to the mile. The aerial mapping program being proposed would facilitate this work.

5.41 It ts moreover desirable that the initial photography of the whole of East Pakistan be done in one season by a specialized aerial photography firm to establish a "bench-mark" year of quality mapping photo cover. Such a procedure would assure the availability of personnel and equipment adequate to meet the needs and the time schedule being projected for the East Pakistan development program. At the same time, the basis would be provided for future needs including sequential photography for continuing studies, prompt photographic reconnaissance and up-to-date photography for each year's mapping program. Besides providing a suitable aircraft. modern air cameras and other equipment, the program should also include training for Pakistani personnel as required to establish a Dermanent aerial photography and photographic unit based in East Pakistan and operated by Pakistanis. Such a unit could be located in a Government agency.

5.42 As such a proposal would involve serviceR of a foreign survey firm it is recognized that the work must be done compatibly with national security nolicy. Consideration might hp given to meet the needs of both security and development by adopting the following principles:

a. The foreign air survey company be, by nationality and reputationn arreptable to trho Cnuornmonr Af Paleistan

b. A security officer be stationed in Dacca to monitor the operation, with discretionary authority to deal with security nroblemn as they arise;

r. As rniiirped in rontrnllinH nortsritvy an nfficpr or his deputy could travel in the aircraft at all times;

d. Exposed film processed at Survey of Pakistan, Dacca, woiild remain utnder his conntrnl And he wunl d atuthnriz,e release of prints to EPWAPDA;

e. He could interpret and administer security policy in respect to enumerated iteu- of basic envi:ronmenrtal data in East Pakistan. - 92 -

5.43 Some months ago, a UNDP proposal was drafted in East Pakistan thru joint efforts of UNDP, EPWAPDA and the Survey of Pakistan, and a PC-i form was prepared to support this program. However, aerial mapping experts from these agencies, and the General Consultants to EPWAPDA all agree that this program is inadequate to meet the needs for photography and mapping when consideration is given to the vastly expanded agriculture and water resources developments being proposed for East Pakistan.

5.44 As a result of conferences held in Dacca during April-May, it was agreed by representatives of UNDP, EPWAPDA, and The Survey of Pakistan that a new PC-1 form and application to UNDP should be drafted.

5.45 The aerial photograph and mapping program should proceed at the earliest possible date. This program will consist basically of the following:

a. The rephotography of the entire Province of East Pakistan at medium scale, principally for air photo interpretation purposes;

b. The photography at large scale of each area scheduled to be mapped at 16 inches to the mile for project purposes in conformity with the mapping schedule;

c. The supply of modern equipment and training of East Pakistan personnel to form an aerial photographic unit capable of meeting future requirements.

5.46 The estimated cost of the aerial mapping program over a 4-year period, as was indicated in Table 3.2 of Volume I, is $3.0 million total equivalent of which $1.6 million is in foreign exchange.

5.47 Geotechnical explorations: Soils testing and analyses used in design of prolect works is carried out by the materials testing section of the hydraulic research laboratory. A considerable increase in capacity will be needed to meet anticipated requirements. Training of personnel in modern techniques is essential, and additional equipment will be needed in the laboratory and in the field. Training of personnel should be accomplished in East Pakistan to a maximum extent. Hlowever, some engineers will require additional training in laboratories outside East Pakistan. Costs were indicated in Volume I, Table 3.2 ($800,000 total eauivalent of which $600,000 is in foreign exchange).

5.48 lnnstmrution materials: Lack of suitable construction materials for river training works and control structures is one of the ma- 4-, probl,ma fai-in,, the wtejrpr developnnnr nrnogram PresntIun Iahbge information suggests that concrete aggregates are principally found in North Bengal and along the Sylhet-Assam border. Supplies of suitable - 93 -

sand are also vailable at nmiber of locarntio%na throiugichouit tha Province. Apart from outcrops of near-surface bedrock in the Chittagong Hill Tracts,

the onLy known potent i a" sources of rock 4n Etas Pakistan are at a d e on the order of 600 feet in the Parbatipur-Rangpur area and even deeper itn tLLLhecoa'L and ±l.Lmestone depost regor.ion of J.raa.lga.j in thne uBogna District.

5.49 An organized exploration for construction materials and th-irough consideration of tihe processiLng, trantspor;, adiu costs ofl provU- ing materials at the job sites should be undertaken soon so that the knowledge gained can be made use of in thiie expanded development program being projected for East Pakistan. Information obtained should be of material value to the toLal construction industry in the Province. Studies would cost $600,000 total equivalent ($300,000 foreign exchange) as was indicated in Volume i, Table 3.2.

5.50 Seismology: East Pakistan is situated in an active tectonic zone. Limited data indicates that a large number of shallow and intermediate depth tremors have been felt within the Provinciai boundaries since seismic data were first collected in the Indian sub-continent around the turn of the century. Historical evidence attests to tne tremendous influence of the tectonic activity in the area with events like the avulsion of the Brahmaputra into the Jamuna some 200 years ago believed to have been triggered by an earthquake, and the tremendous landslide into the Brahmaputra Valley in Assam in 1950 which has caused considerable aggradation of the bed of the river and is probably still affecting the behavior of the Jamuna in East Pakistan. Large hydraulic structures must be designed to withstand earthquake forces, and without adequate pertinent information, overdesign for safety can result in increased costs. To provide rational criteria for design, it is proposed to obtain expatriate advice on analyses of data and the procurement of seismographs for stra- tegic location in the Province. Presently, the only seismograph location is in Chittagong and it is understood that the data have not been fully analyzed. Study costs are estimated at $200,000 total equivalent ($100,000 foreign exchange) as was indicated in Volume I, Table 3.2.

5.51 Geomorphology and sedimentology: East Pakistan is traversed by large rivers carrying heavy loads of sediment. These rivers are constantly modifying their course and changing their cross section. Successful river training for flood control and construction of water regulating facilities is dependent upon properly identifying these conditions utilizing up-to-date techniques of analysis in the field of river morphology, sedimentology and fluvial hydraulics. While much information is already available from specific project studies (e.g. Dacca Southwest) and further data and advice will be available from the flood model studies (para. 5.23) and from the Flood Consulting Board (para. 5.30), basic research and long-term training of Pakistani personnel should be started at an early date. - 94 -

5.52 Several universities have wide experience and staffing expertise to assist Pakistani staff in learning modern methods and techniques of analysis in sedimentology, geomorphology, and fluvial hydraulics. Their laboratories include facilities for testing and analyzing specific field conditions as necessary. These institutions are equipped also to provide educational training. A limited number of Pakistani engineers could receive fundamental training under this program. Costs are estimated at $100,000 total equivalent all of which is foreign exchange; see Volume I, Table 3.2. CALf PAKIS7117-- X07T 0'OFT7AI3 1A(7T7'.1

lype Re-aining (;ot after of 1?70-71 197L-72 1972-73 - 19 72-73 Pro w, Total P R J 0 1 7 T S Immestwsot i.-ak. St.rt complee ~ ~ lca o~g Ttl t'a -.IPr 3ori,gLn TO=i ./TPr~g To/i-. tolPrM EIPWPODAPROJECTS

1. Gss.mge-tobadack, Kashtia Unit Part (a) Capital Inm. 1/7/70o 30/6,'72 7.5 2 .5 3.0 0. 7 2.-5 3.0 1.7) Part 5.0 6.0O (b) Capital Inm. 1/7/73 30/6/75 19.0 7.0 26.0 19.71 7.0 2605 Milt. A.et. 1/7/71 30/65/75 Q1. 0.1 0.6 0.11 0.1 !Q- 0.7 0.1 !)-4 0.6 0.2 17.8 j,'5 t, 2.0 .2.8R 2.6 3.7, 0.I9i 2.6 3.7, 0.3 0.1 0.L 19.6 7.2 26.8 21 .5' 12.5 A ,.o 2 . Coastal Ebsakat Ph,Ase I - 0oop A Capital In-. 1/7/70 1/65/72 2 9.0 1.0 30.0) 9.0 1.0 10. 0 38.2 2.0 40. 0 tar.p A (Water kt.) Proj. Prep. 1/1/11 30/5/774 0.2 0.l 0.13 0.3 0.1 0.74 0.3 0.1 0.74 0.2 Gmoap B 0.1 0.3 1.0 0.74 1.74 Proj. Prep. 1/7/ 70 30/65/72 03.2 0.1 0.3 2.2 0.6 .3.0 2.1, 0.9 :1.3 Group B -Stage I Capital In-. 1/9/,72 33/16/75 3.13 1.2 5,.0 10.1o 3 .3 13. 3 13.81 Oroup B Stage 74.5 113.3 11 Capital Imm- 1/7/15 31/12/78 13.1 74.6 18.74 13.11 74.6 113.4 Phs.e II - Projeot Preparation Proj. Prep. 1/1/71 30/16/774 C.1I 0.1 0. 2 0. 3 0.2 0.5 0..)~ 0.2 77.5 0.13 0.1 02.3 0.-9 0.6 :1.5 Stage I Capital 1mm- 1/9/172 30/6/75 74.0o 74.0 13.0 16.70 6.0 223.0 20.7) 10.0 372.0 Stag U1 Capital In,. 1/9/75 30/65/78 20.1) 10.0 377.0 20.7) 10.0 372.0 Ph,aaee I &rl Overall Mlgt. Anot. - Pert I 1/7/71 30/65/774 0.1 0,1 72.2 1.0 0.5 :1.5 1.74, 0.7 3.1 2.5; 1.3 3).8 Part II 117717h 31/1:3/77 -- - -2 .6 IIj J-9 2.6i L. .L2 39.5 1,3 30.73 11. 9 2.2 14.1 9.1, 6.0 15.74 674.2 26.1 90.3 11,5.02 35.6 150.6 3. ChAodporz Capital Imm- IEA Credit 1/1/70 30/6/73 73.0 2.6 10.6i 7.5 2.0 9. 5 0.73 0.74 77207 .,1. . 1. 74. fleoa Sc.otweat Capital Cm-. 1/1/71 30/65/76 .3.0 1.0 74.o 8.-5 2.5 11.0 9.5 74.5 11s.0 274.0) 7.0 31. 0 745.72 15.0 60.0 Final Emg. )1, ,)A CrediLt 1/1/170 31/12/70 0.6 0.8 1.7, 0.6i 0.8 :1.74 Proj. Prep.)

3.6 1.8 5.7Is 8.5 2.5 11. 0 9.; 74.5 lTh.o 24.7) 31.0 31L. 745.6 15.8 6:.1, 5. arnafoli Capital In-. CIA Credit 111/72 1/l/75 0.7 0.3 :1.0 7 .5 2.5 177.0 6.73 3.2 10.0 15.72 6.0 21. 0 Ploa l g. ) 1/12/70o 1/13/71 72.5 1.0 1.5 0.5 1.0 1. 5 o.5 0.74 77.9 Proj. Prep.v 1/7/70 1/5;/71 1.5 2.74 :3.9

77.5 1.0 1.5, 1.2 1.3 2.5 8.0 2.9 177.9 6.33I 3.2 10.0 16.5 8.4, 24,.9 6. mubutri Capital CIm. Iwluded 1/1/?3 1/:1/76 0.7 0,3 1. 0 7.5 3.5 11.0 8.13 3.8 13.0 Ftaal &g. int 1/8/71 1/123/72 Proj. Prep. Karoafuali 1/2/71 1/74/72 -

0.7? 0.3 1. 0 7.5 3,.5 11.0 8.13 3.8 U1.0 7. Thbml7.e I Thakargamo ttenseion Capital lay. 1/7/71S 1/6/74 0.7 1,9 27.6 1.1 3,8 5,.0 1.3 3.8 5.0 3.11 Tsutulta P.S U1.6 Parnohagarh Capital Lam. 1/7/71 1/6/75 0.13 1,1 1..74 2.7) 6,0 33.0 3.7) 9.10 11.0 5.3 16.1 21. 4 tanIgeu C-apital Lwm. 1/7/171 1/6/774 0. 2 0.74 77.6 1.5 745 6.0O 1.5; M.rt1s 74.5 65.0 3.13 9.74 113.6 M3nessoiog 1/7/112 1/67/75 0.5' 1.5 17.0 2.77 6.0 33.0 2.5 7.5 10.0 P'Loal E1g. & Al-l Tube- 1/7/10 30/6/72 . P7mj. Planmiog teells I7.74 0.6 1.77 0.3 0.7 1..0 -. -- - - - 1 11.1 13.0 Tot.el Toboumlla I 17.. 0.6 1.17 1.5; 6.1 5. 6 5.17 15.5 21..3 7.7~ 23..3 311.0OoJ,.3 743.8 533.6 8. Pabma Phaise I Capita I.m. 1/9/7j2 1/7'/76 L5 15 603- L s03. 25 5. Final tng. 1/1/1-' 30/6/72 0).5 1.5 2.07 0.5 1.5 2. 0 1.7) 3.0 Is. 0 Proj. Prep. Completed iss .Ld-1970 ------

(1.5 1.5 2.0C 0.5 1.5 2.0 L4.5 1.5 6.0 33.73 111.0 744.0 38.5 15.5 574.0 9 . Bm7lkuhi aepital too 7/7/72 37/6/75 3.77 1.0 74.0 15.17) 5.0 277.0 18.77 6.0 23,.0 R,L.al Erg. ) IncIuded L. P- J. PreT1) Puabn

3.77 1.0 ,.O 15.17 5.0 217.1 18.77 6.0 21,.0 I r, 14+rleT W-i P-- 7,','7 J.'~117/3i ).5 r7.29 22.5; 7.5 317.0 22.5 7.5 3s .0 9>ial oEt. 1/7/172 31/11/73 7. 7 7 . 2/ 9 .l~2/o,' 0.2,' 1. 1. P,~~~o1 Pr-r 1,1.77~~~~~~~3t,T'1,'7 7. , 7 .5 o. 4/ 1.7 .5 1. ?/ 7.77 1.0 1. 8

11. 7~'Per Ku~7-o, .',o71. J''?." . 3.7 7 . 5.5; 5 7 5 17.77 3,5 S T '~~~17713-7 /,'72 .3 I71.'. O.07 0.9 L-5

-...... ~~~~ ~~~~~~~~~~~~~~2.7'.5 10.17 74.4 115.0 OaaloSt After Type o f 1,72-71 i `7 I-'r'2) :2 -> 1772 -1- PrE TaiaI

IV. Koah..ao o. bob--t 1,1"72 1, 1, 75 2.? U. 3 It li.21 2.9 56. 2. 2 6.3 . 3.1 11.0 F-ca king- I :o, 077cr 1/7,71 30'l Proj. Prep.- ya,KIoblyara 117 1.3 10 1 .4. <. > : . . . 3.2 22.3

133. Dacca Narth Capital Ice. L/7/73 1/7/77 3.9 19.0 42.3 32.,) 1011. 13.11 FinalBrg. 1! 7/72 31/12/73 0.1 2L.5 3.7 O. 1.1 1.7 3.p 1.7 2.5 Proj. Pc-p. 'L/1I31I 43,16/72 D., 0.5 13.7 0.3' 0-5 0. 8 . . 1.5 1.2 0.5 0.7 .S 1.0 1.6 33.6i 11.1 Ui.? 31.6 12 .6 46.30 13-.galiC ?ha.e I Project frn1p.rat.1an 2/1/71 30/6/73 0.3 o.7 1.0 0.5 1.0 1.5 9.5 1.0 1.51 1,3 2.2 4.0 Final Kagiarerling Far .11 stages [/1/71 30/6/73 0.3 0.7 1.0 0.2, 1.5 2.0' L. .5 1.5 2.0 1.3 3.7, 5.0 Staire (a) CapItal Ian- 0/1/71 33<,6/74 1.5; 0.5 2.0C 2.3 0.7 3.0 0.7 0). 3 1.0 L.5 1. 5 6.o Stlag. (b1 Capital In-. 1/1/73 )L3/6/76 0/ (.3 1,0 10.2 3.0 13.0 10.7 3.: IL 0 Stags". Capital Inn. L/1/74 37:/6,'773 12.92 4.0 16o 1-2. 40 16,0 0.6 i.6 2.0 2.5 3.0 5.5 4.0 j).5 73.5 22.7 7.3 32.0 29.5 15.-2 4550

16. aal-N&hl. Phase I Capital Ian- '1/12/73 30/6/77 i2/ 3 5 1. 00 3. 2!' 5. Finaltog. 7/7/'72 431/3 . . 1.5 0.7 1.3 20 0.5 1. 15 1.7 3. 50 Proj . Prep. 1.'1/72 33/6/73 0.3 0).7 1.0) 0.3 0.7 1.0 04 0. 7 1.0 0.9 2.:. ~.O. o.6 1.7 2. 5 1.9 0.11 3.0~ 35.3 14.2 52.6 Lo.v 17.5 58.0

Io. Sanog: Capital Inv. L/1/35 1/7/77 37 .5 12.5 5o.o 37.5 12.5 50.13 Pin-I fisg. ~)1/7/72 31/02/73 0.3 1.0 1. 3 0.39 .0 3.3 o4 1.0 1.4 1.0 3. 4.0

3.3' 1.0 1. 3 0.3) 1.0 1.3 37 .9 13.5 51.4 35.5 15.5 54.0 17. Tube.l1o II Iafltal Inv. 1/1/74 1/1/78 1.0 2.o 3.13 9.1 29.5 39.0 10.5 32-5 42.0 ftIn. 1a.6 1/1/72 33/6/73 0.1I 0).2 0. 3 0.2 13.2 0.1 0.4 0.5 P'roj, PrM. 1/12/70 32)/6/72 0.2 0I.3 0.5 .6 0.6 1.06 _0.91 L_ 0.2 0).3 0-5 0.5 3.8 1.3 1.0 2.2 3.2 9.5 29.5 35.0 11.2 32.15 44.0 19i. fagei,-Jnb.dakc, Je...are Unit (.Phae 015) Proj. Prep. 1/6/73 32/6,74 __ 0.5 1,,,,j 2.0 0. I..' 2.0 Tot-el SPVAFDAProject, cj, 31 57.2 39.-1 25.3 64.6 53, 1'5.4 99.3 351.7175.7 528.4 490.0 259/.5 74p.5 flood onItrol inveatiationsJ 1. flnges, Barra.ge and Soathveet Regio, 1/ 1/72 11112/7, 5. 1.5 2.0 o.-5 1.5 2.0 9,5 1. i 1 .5 1.-5 4.1o 5.5 2. Mbodl of Sot-we-t Reag:o. 1/7/71 30/86/79 0.3 0.3 1.0 0.3 0.7 0.9) 0.6 1.4 2.0 3. Flaco K-ce11 1'7 r7 1 3V69/76 u.Sl 1.0 1.5 0.3 0.7 1.9 0,9 1.7 2.5 a. Bralapetre - Oecges-Keghno C~pjev 1/1/7 33/6/71 0. 1.4 2.0 <1.6 i .4 2.3 1.3c A .7 1,0 1.5 3.5 5.0 Total FL-u Control ±noe-tjgatlr- 1.2: 3.6 5.0 0.9Y 4.6 6.5 2.1 2.1t 3.5 .4 10.15 15.0 A00 Parajlao

1.Lo.11/it hoops I 1/1/71 31,6/73 3.1 16. 7. 4.11 3.1 7. O 7.0 7.3 14.0 2. 1-l,Vr ?=o1 1/ 13/3 92,<6/75 6.0 2.o 5 .2, 16.0 3.0 137.0 20.0 5.2 3)5.0 3. TafreWallsI 1<12/73 31/12/73 3. $.6 .1. 6.5 8.9 15.i, 6.5 4.9 11.4 5.3 3.2 6L~.7 29.9 22.9 64.6 L. Tubevelh 11 1,02/73 34131767 -. - - - 6 6.3 1. -.1' 35-1i 7t.4a 60.0 4!2.13 a4.0 Total A31CPrjojeto .6 .6 16.1. 13. 11.9 22.6, i9.5 13.0 32.0 953 42.1 117.1 90.9 76. 167.6

Orcad Total j%(6 7/? <3. 60. 405 .7-) 76 37 137. 1.13 2373. 623.13 586.2 346.9 2.~2-1

2: .

., Boricoct00! :00~1a:t,:: nv ,fe.o o 0,L :: 1.ac) oa:vit Table 2

ARFAS I?RTIA?E/V AND 'RON narr.nfl.J)ODIf TtSeaoLnd or aceT)-

1975-76 1980-1 i05i-86 - mitiate _Rerks .oed Flood[ ni=-OCTod ProtcUtion Irrigatiol Protection Irrigation Protection Iri1gation Protection IrrLgstion

LB-1uft F-py 1,000 _0- 500 - o 500 Peek 1,100 in 1976-77 tbemelis (-on-ints tve)- 350 6C10 - 300 *D 300 Pszk 8W0 in 1978-79

bsel (ttendrew - 200 - 1,2,200 2t,ZS00

M14ti-purpose Proj*acw NowProjects NC-Ca Sauteet 260 130 39o 260 390 300 390 30 Karma.uli 100 20 150 70 no 100 150 1W00 Nuhuri 50 10 lC1) 5i0 100 60 100 60 Pabna 1 220 40 220 160 220 160 220 1!6o Patr 11 _ - 230 Eo 230 120 230 160 Bel1icJhi 1OD 30 100 '70 100 70 100 70 Upper Kustyara 80 70 Do 110 80 50 130 50 Yi.oal 60 20 80 L,o 80 50 50 50 Sangu 20 10 120 EO 120 80 120 80 Dacca !'orth - - 161 100) 160 110 1/50 110 Barthal t _ 50 30x) 160 300 200 300 .200 BarlF%1 rl - - 150 lrc 150 310 900 600 Little Fenri I 141 80 111 90 1-1lz0 90 Little Fena II - 120 LO 120 70 120 70 Cordl1a-;oako.a1L I - - 220 * 160 220 160 220 16O -Co'fl)a-toak li II - - 3°° 1iQ 670 3110 6Syo 15

Subtotal, NaW Projxct. 910 330 2,60) 1,5950 3,230 2,300 3,9130 3,71o

.int!ng_Projects ia !ikentm 1,500 350 1,800 1,100 2,4100 1,600 3,0t00 2,500 GKi-.Pjsbhti aD' 350 130 350 220 350 220 350 ;.20 Chanrtpar ihO 60 140 910 lhO 90 11I0 _ 9

Subtotal, ExLettng Project 1,,990 50 2,290 1,41LO 2,B90 1,910 3,110 2,810

Total, Mul1ti-prpose Projects 2,900 870 5,150 3,QX0 6,120 ,210 7,170 6,1;20

G?.CD I?TAL 2,,r900 2,420 5,15Q 5,100 6,120 7,210 7,1170 9,5'20

I/ !et area cormanded. Not including existing (1969-70) area of 70,CO acrets.

oGross area protected. Not itncluding existitng DV6e-Dest project (15,000 acres g]rss, 10,000 acres net) Not including existiitig (1959-70) area (of 700,000 ac-res. m' Srahuzputra Right Ebnkm.ent flood p.rotectLon proJect 1(580,00 acres gross).

Not including existinj (1-959-70) area okf 4o,ooo acres. Includes 10,000 acres now partially irrigated.

Table .3

:ICEN0BfMUAL AAGRI'CULTURAL PRODUCTION FROM WATER DEVELOP]MENT PROJECTS

7L/72 72/73 7.3/74 7475 75/76 76/77 77/78 78/ 79 /80 80/81. 85/86

- REA IRRIGATE]D (ihousandsi of acres) 500 Low-lift pumps 300 500 700 90C0 10CI 110C) 90g) 700 500 500 300 Tibewell (non-intensive) 50 100 150 250 350 50C) 70( 800 700 600 10(D 2200 rubewells (inatensive) - 15 70 1140 22C) 27' 405 56 5 76D 4210 Xulti-pmrpose projects 40 150 300 540 870 113() 1820 2330 2770 3000

Totals 390 750 1220 1830 2440) 3005 3825 4395 4730 5100 7210

- AREA EFFECTLVEL! IRIGATED (thousands of acres)

Low-lift pumps 120 320 51c) 69() 83( 900 871D 720 540 460 1460 Tubewells (non intensivei) 20 5c) 10C) 160 240 350 480 620 680 580 27'0 2000 Tubewells (intensive) - - 20 50 110 210 330 470 61.0 760 iWlti-purpose projects IC) 40 160 300 420 _ 1170 1590 205Q0 2370 3800 Totals 150 410 790 1200 1600 2240D 2850 3400 392.0 417'0 6530

. - ;ARCULRURALT PRODJIJCTIUI (thousanids of long tons) Low- lift. punips 114( 38( 610 830 990 1040 1020 860 650 55io 550 Tubewell.s (non-intensive) 40 100 100 320 1480 700 960 1240 1360 1160 5140 Tabewells (intensive) - - 50 125 275 525 830 1180 1520 1900 5000 Mlti-purpose projects 30 120 480 900 1260 2350 3500 4800 62550 7100 11400 Totals 210 600 13140 2175 3005 4615 631.0 8080 9780 10710 17490

MAP I | NEPAL ! j4-t B\H U T AAN 7EATPKS N I / ta t\ X1 T-i , ' P~~~~~~~DRESENr\TILANDn ISlr[

| nt . jx aS gB i-a A L t 'o {' j z~~~~~EET icdals wamnnps- S urnd a rbrn foarest

|/-<(-Xs j | S\ l' < lt Ji 0 r cl ~~~~~~~~~~Lowlard- broadcast Amen rice and j.te

;2%GE5n_ \} # ^S,<54 /tS J * | X t t{S\l < WXV-r lowland- 9a- rice

-26° #) 5 I ! (t ( J L ~~~~~~~~~~~~~~~~~~~H,gh land (old alluvium)-fransplanted Amon r,ce q OIC| P| tUr L Inemdrelndv,,obie Gind !,ip uerppng i s . -:;X s i /7 _I e ,,, ~~~~~~~~~~~~~~land-fea, fruits and forest | D.t Vjjnlnr\eDi S G ;)4 }

I \,/l/>nAX Cts1's.~N-> t,- v.ASSAM-

1'<-)~~~~~~ o c>-. 4, Ix -I.AhPf

n _/ mA~~~~~~~~~~t;uVl, - t f"za l~~~~~~k~ "I l: _l _.

2 4 AITN _ <

JUNE 1970 IBRD-3026~~b1

MhAAP2

a80~~~~~ UT A N 9 oo EAST PAKISTAN NPAL ' UTAN _ WATER PROJECTS i/ 1'' NXD t1 A '-AGRICULTURE & WATER DEV'ELOPMENT

oj\ J / 9 \ 1 t > / ) ( > / t / ~~~~~~~~~~~~~~~~~~~BrahmrGp,,t, right em,bonkm,-e,,l

X W& p N ; < - - T:;OV ' bou~~~~~~~~~~~~-----ndor,e 6° \ w)1X > \\+ s / d}J roetrues

h 0n 6i l t\ sh"n 1 eFn

) / l\ / 1} (7'8 o2X1/5 t. g 9.B~~IGanes-K ob.i da,Kstoui 18 LittlesFe,, o,Jesreui A/// 1; Jfi~~~~~~~~~~~~~~~~~~CasaEt- S emanmet 11X'Uppe,r Kusi - I

MARyXf~~~~~~~~~~~~~~~~~A 4Dcc Sot.s D 13. DaccagoNoMXes

P~~~~~~~Bayof BengaM 'lsO 1lhi r-q-fi N-0hl,;

tJ 91°- KILOMET~~~~~MER92-'. i JULY1970 i9RD 3027R~~~~~~~~1